[libvirt] RFC: API additions for enhanced snapshot support
Right now, libvirt has a snapshot API via virDomainSnapshotCreateXML, but for qemu domains, it only works if all the guest disk images are qcow2, and qemu rather than libvirt does all the work. However, it has a couple of drawbacks: it is inherently tied to domains (there is no way to manage snapshots of storage volumes not tied to a domain, even though libvirt does that for qcow2 images associated with offline qemu domains by using the qemu-img application). And it necessarily operates on all of the images associated with a domain in parallel - if any disk image is not qcow2, the snapshot fails, and there is no way to select a subset of disks to save. However, it works on both active (disk and memory state) and inactive domains (just disk state). Upstream qemu is developing a 'live snapshot' feature, which allows the creation of a snapshot without the current downtime of several seconds required by the current 'savevm' monitor command, as well as means for controlling applications (libvirt) to request that qemu pause I/O to a particular disk, then externally perform a snapshot, then tell qemu to resume I/O (perhaps on a different file name or fd from the host, but with no change to the contents seen by the guest). Eventually, these changes will make it possible for libvirt to create fast snapshots of LVM partitions or btrfs files for guest disk images, as well as to select which disks are saved in a snapshot (that is, save a crash-consistent state of a subset of disks, without the corresponding RAM state, rather than making a full system restore point); the latter would work best with guest cooperation to quiesce disks before qemu pauses I/O to that disk, but that is an orthogonal enhancement. However, my first goal with API enhancements is to merely prove that libvirt can manage a live snapshot by using qemu-img on a qcow2 image rather than the current 'savevm' approach of qemu doing all the work. Additionally, libvirt provides the virDomainSave command, which saves just the state of the domain's memory, and stops the guest. A crude libvirt-only snapshot could thus already be done by using virDomainSave, then externally doing a snapshot of all disk images associated with the domain by using virStorageVol APIs, except that such APIs don't yet exist. Additionally, virDomainSave has no flags argument, so there is no way to request that the guest be resumed after the snapshot completes. Right now, I'm proposing the addition of virDomainSaveFlags, along with a series of virStorageVolSnapshot* APIs that mirror the virDomainSnapshot* APIs. This would mean adding: /* Opaque type to manage a snapshot of a single storage volume. */ typedef virStorageVolSnapshotPtr; /* Create a snapshot of a storage volume. XML is optional, if non-NULL, it would be a new top-level element <volsnapshot> which is similar to the top-level <domainsnapshot> for virDomainSnapshotCreateXML, to specify name and description. Flags is 0 for now. */ virStorageVolSnapshotPtr virDomainSnapshotCreateXML(virStorageVolPtr vol, const char *xml, unsigned int flags); [For qcow2, this would be implemented with 'qemu-img snapshot -c', similar to what virDomainSnapshotXML already does on inactive domains. Later, we can add LVM and btrfs support, or even allow full file copies of any file type. Also in the future, we could enhance XML to take a new element that describes a relationship between the name of the original and of the snapshot, in the case where a new filename has to be created to complete the snapshot process.] /* Probe if vol has snapshots. 1 if true, 0 if false, -1 on error. Flags is 0 for now. */ int virStorageVolHasCurrentSnapshot(virStorageVolPtr vol, unsigned int flags); [For qcow2 images, snapshots can be contained within the same file and managed with qemu-img -l, but for other formats, this may mean that libvirt has to start managing externally saved data associated with the storage pool that associates snapshots with filenames. In fact, even for qcow2 it might be useful to support creation of new files backed by the previous snapshot rather than cramming multiple snapshots in one file, so we may have a use for flags to filter out the presence of single-file vs. multiple-file snapshot setups.] /* Revert a volume back to the state of a snapshot, returning 0 on success. Flags is 0 for now. */ int virStorageVolRevertToSnapsot(virStorageVolSnapshotPtr snapshot, unsigned int flags); [For qcow2, this would involve qemu-img snapshot -a. Here, a useful flag might be whether to delete any changes made after the point of the snapshot; virDomainRevertToSnapshot should probably honor the same type of flag.] /* Return the most recent snapshot of a volume, if one exists, or NULL on failure. Flags is 0 for now. */ virStorageVolSnapshotPtr virStorageVolSnapshotCurrent(virStorageVolPtr vol, unsigned int flags); /* Delete the storage associated with a snapshot (although the opaque snapshot object must still be independently freed). If flags is 0, any child snapshots based off of this one are rebased onto the parent; if flags is VIR_STORAGE_VOL_SNAPSHOT_DELETE_CHILDREN , then any child snapshots based off of this one are also deleted. */ int virStorageVolSnapshotDelete(virStorageVolSnapshotPtr snapshot, unsigned int flags); [For qcow2, this would involve qemu-img snapshot -d. For multiple-file snapshots, this would also involve qemu-img commit.] /* Free the object returned by virStorageVolSnapshot{Current,CreateXML,LookupByName}. The storage snapshot associated with this object still exists, if it has not been deleted by virStorageVolSnapshotDelete. */ int virStorageVolSnapshotFree(virStorageVolSnapshotPtr snapshot); /* Return the <volsnapshot> XML details about this snapshot object. Flags is 0 for now. */ int virStorageVolSnapshotGetXMLDesc(virStorageVolSnapshotPtr snapshot, unsigned int flags); /* Return the names of all snapshots associated with this volume, using len from virStorageVolSnapshotLen. Flags is 0 for now. */ int virStorageVolSnapshotListNames(virStorageVolPtr vol, char **names, int nameslen, unsigned int flags); [For qcow2, this involves qemu-img -l. Additionally, if virStorageVolHasCurrentSnapshot learns to filter on in-file vs. multi-file snapshots, then the same flags would apply here.] /* Get the opaque object tied to a snapshot name. Flags is 0 for now. */ virStorageVolSnapshotPtr virStorageVolSnapshotLookupByName(virStorageVolPtr vol, const char *name, unsigned int flags); /* Determine how many snapshots are tied to a volume, or -1 on error. Flags is 0 for now. */ int virStorageVolSnapshotNum(virStorageVolPtr vol, unsigned int flags); [Same flags as for virStorageVolSnapshotListNames.] /* Save a domain into the file 'to' with additional actions. If flags is 0, then xml is ignored, and this is like virDomainSave. If flags includes VIR_DOMAIN_SAVE_DISKS, then all of the associated disk images are also snapshotted, as if by virStorageVolSnapshotCreateXML; the xml argument is optional, but if present, it should be a <domainsnapshot> element with <disk> sub-elements for directions on each disk that needs a non-empty xml argument for proper volume snapshot creation. If flags includes VIR_DOMAIN_SAVE_RESUME, then the guest is resumed after the offline snapshot is complete (note that VIR_DOMAIN_SAVE_RESUME without VIR_DOMAIN_SAVE_DISKS makes little sense, as a saved state file is rendered useless if the disk images are modified before it is resumed). If flags includes VIR_DOMAIN_SAVE_QUIESCE, this requests that a guest agent quiesce disk state before the saved state file is created. */ int virDomainSaveFlags(virDomainPtr domain, const char *to, const char *xml, unsigned int flags); Also, the existing virDomainSnapshotCreateXML can be made more powerful by adding new flags and enhancing the existing XML for <domainsnapshot>. When flags is 0, the current behavior of saving memory state alongside all disks (for running domains, via savevm) or just snapshotting all disks with default settings (for offline domains, via qemu-img) is kept. If flags includes VIR_DOMAIN_SNAPSHOT_LIVE, then the guest must be running, and the new monitor commands for live snapshots are used. If flags includes VIR_DOMAIN_SNAPSHOT_DISKS_ONLY, then only the disks are snapshotted (on a running guest, this generally means they will only be crash-consistent, and will need an fsck before that disk state can be remounted), but it will shave off time by not saving memory. If flags includes VIR_DOMAIN_SNAPSHOT_QUIESCE, then this will additionally request that a guest agent quiesce disk state before the live snapshot is taken (increasing the likelihood of a stable disk, rather than a crash-consistent disk; but it requires cooperation from the guest so it is no more reliable than memballoon changes). As for the XML changes, it makes sense to snapshot just a subset of disks when you only care about crash-consistent state or if you can rely on a guest agent to quiesce the subset of disk(s) you care about, so the existing <domainsnapshot> element needs a new optional subelement to control which disks are snapshotted; additionally, this subelement will be useful for disk image formats that require additional complexity (such as a secondary file name, rather than the inline snapshot feature of qcow2). I'm envisioning something like the following: <domainsnapshot> <name>whatever</name> <disk name='/path/to/image1' snapshot='no'/> <disk name='/path/to/image2'> <volsnapshot>...</volsnapshot> </disk> </domainsnapshot> where there can be up to as many <disk> elements as there are disk <devices> in the domain xml; any domain disk not listed is given default treatment. The name attribute of <disk> is mandatory, in order to match this disk element to one of the domain disks. The snapshot='yes|no' attribute is optional, defaulting to yes, in order to skip a particular disk. The <volsnapshot> subelement is optional, but if present, it would be the same XML as is provided to the virStorageVolSnapshotCreateXML. [And since my first phase of implementation will be focused on inline qcow2 snapshots, I don't yet know what that XML will need to contain for any other type of snapshots, such as mapping out how the snapshot backing file will be named in relation to the possibly new live file.] Any feedback on this approach? Any other APIs that would be useful to add? I'd like to get all the new APIs in place for 0.9.3 with minimal qcow2 functionality, then use the time before 0.9.4 to further enhance the APIs to cover more snapshot cases but without having to add any new APIs. -- Eric Blake eblake@redhat.com +1-801-349-2682 Libvirt virtualization library http://libvirt.org
Hi, I'm new to this ML, and very interested in this topic. I'm going to try to implement LVM snapshot feature using this framework. 2011/6/16 Eric Blake <eblake@redhat.com>
/* Opaque type to manage a snapshot of a single storage volume. */ typedef virStorageVolSnapshotPtr;
I'm just wondering how to detect storage type and to choose assosiate shapshot functionality. Is there any idea about it?
/* Revert a volume back to the state of a snapshot, returning 0 on success. Flags is 0 for now. */ int virStorageVolRevertToSnapsot(virStorageVolSnapshotPtr snapshot, unsigned int flags); [For qcow2, this would involve qemu-img snapshot -a. Here, a useful flag might be whether to delete any changes made after the point of the snapshot; virDomainRevertToSnapshot should probably honor the same type of flag.]
For LVM, it may be lvconvert --merge , but requires recent version of lvm and linux kernel, So, there are some failure reason (toolchain doesn't support it or toolchain supports, but failed) I don't know how to handle these case for now, but we may need to add some virErrorNumber to indicate snapshot-related error. - 元のメッセージを表示 -
As for the XML changes, it makes sense to snapshot just a subset of disks when you only care about crash-consistent state or if you can rely on a guest agent to quiesce the subset of disk(s) you care about, so the existing <domainsnapshot> element needs a new optional subelement to control which disks are snapshotted; additionally, this subelement will be useful for disk image formats that require additional complexity (such as a secondary file name, rather than the inline snapshot feature of qcow2). I'm envisioning something like the following:
<domainsnapshot> <name>whatever</name> <disk name='/path/to/image1' snapshot='no'/> <disk name='/path/to/image2'> <volsnapshot>...</volsnapshot> </disk> </domainsnapshot>
where there can be up to as many <disk> elements as there are disk <devices> in the domain xml; any domain disk not listed is given default treatment. The name attribute of <disk> is mandatory, in order to match this disk element to one of the domain disks. The snapshot='yes|no' attribute is optional, defaulting to yes, in order to skip a particular disk. The <volsnapshot> subelement is optional, but if present, it would be the same XML as is provided to the virStorageVolSnapshotCreateXML. [And since my first phase of implementation will be focused on inline qcow2 snapshots, I don't yet know what that XML will need to contain for any other type of snapshots, such as mapping out how the snapshot backing file will be named in relation to the possibly new live file.]
In LVM case, taking snapshot(lvcreate) needs at least snapshot volume size as an argument, So, I think storage-specific element can be inserted in this XML format
Any feedback on this approach? Any other APIs that would be useful to add? I'd like to get all the new APIs in place for 0.9.3 with minimal qcow2 functionality, then use the time before 0.9.4 to further enhance the APIs to cover more snapshot cases but without having to add any new APIs.
LVM snapshot may become invalid in the case of running out the volume size, So, adding an API for checking whether snapshot is valid will be needed. Regards,
On 06/20/2011 04:22 PM, Motonobu Ichimura wrote:
/* Opaque type to manage a snapshot of a single storage volume. */ typedef virStorageVolSnapshotPtr;
I'm just wondering how to detect storage type and to choose assosiate shapshot functionality. Is there any idea about it?
I'm not yet convinced we need storage-type-specific functionality; the XML elements can be optional, and if a particular element is not needed by a particular storage driver, then that driver can ignore the element (that is, the point of the snapshot XML on the create API is to provide details like name or description where the default is not good enough, and on dumpxml it will be the minimum xml for what was actually used in creating the snapshot). I believe that libvirt already has code for distinguishing between different storage types (directory vs. block device), and can therefore already tell whether a block device is an LVM partition. It also shouldn't be too hard to tell whether a disk image belongs to a btrfs mount point, for using btrfs copy-on-write cloning (if nothing else, coreutils' cp has already done this, so I have examples to copy from). But I'll probably have to get more familiar with that code as I expand this API beyond qcow2, and it certainly makes sense to consider whether 'virsh vol-info' should be expanded to display information already known by libvirt (whether the volume is file- or block-based; if block-based whether the partition is LVM, and if file-based what file system the containing mount point appears to be). That may indeed involve adding more APIs beyond this proposal, but I think that this proposal will work for 0.9.3 even if such new APIs are delayed until after the release.
/* Revert a volume back to the state of a snapshot, returning 0 on success. Flags is 0 for now. */ int virStorageVolRevertToSnapsot(virStorageVolSnapshotPtr snapshot, unsigned int flags); [For qcow2, this would involve qemu-img snapshot -a. Here, a useful flag might be whether to delete any changes made after the point of the snapshot; virDomainRevertToSnapshot should probably honor the same type of flag.]
For LVM, it may be lvconvert --merge , but requires recent version of lvm and linux kernel, So, there are some failure reason (toolchain doesn't support it or toolchain supports, but failed) I don't know how to handle these case for now, but we may need to add some virErrorNumber to indicate snapshot-related error.
Yes, we can add new virError values as we come across situations where such operations are unsupported, but which don't fit well into any existing error values.
virStorageVolSnapshotCreateXML. [And since my first phase of implementation will be focused on inline qcow2 snapshots, I don't yet know what that XML will need to contain for any other type of snapshots, such as mapping out how the snapshot backing file will be named in relation to the possibly new live file.]
In LVM case, taking snapshot(lvcreate) needs at least snapshot volume size as an argument, So, I think storage-specific element can be inserted in this XML format
Wouldn't the snapshot volume size always be equal to the volume size of the original? Or is there really value in allowing the snapshot size to be configured to something different (and if different, must it always be <= original, or does a larger snapshot than the original make sense)?
Any feedback on this approach? Any other APIs that would be useful to add? I'd like to get all the new APIs in place for 0.9.3 with minimal qcow2 functionality, then use the time before 0.9.4 to further enhance the APIs to cover more snapshot cases but without having to add any new APIs.
LVM snapshot may become invalid in the case of running out the volume size, So, adding an API for checking whether snapshot is valid will be needed.
I'm not quite sure I follow the scenario you are envisioning here. Would you mind stepping through an example (mapping [proposed] libvirt API calls to lvm commands would be helpful), on how an original lvm partition is created, then a snapshot partition, then how you run out of volume size in that snapshot? It sounds like you are saying that there is a way to create an lvm snapshot that is valid at the time that is created, but later on, subsequent actions cause the snapshot to run out of space and no longer be valid. But my understanding is that a snapshot is a constant size (it represents a known state of the disk at a fixed point in time), only the deltas to that snapshot (aka the live disk) ever have the potential to grow beyond the amount of storage used by the snapshot. Or are you worried about creating an lvm snapshot by libvirt, but then a third party program changes the property of the lvm snapshot volume to change its size? -- Eric Blake eblake@redhat.com +1-801-349-2682 Libvirt virtualization library http://libvirt.org
On 06/20/2011 04:52 PM, Eric Blake wrote:
LVM snapshot may become invalid in the case of running out the volume size, So, adding an API for checking whether snapshot is valid will be needed.
I'm not quite sure I follow the scenario you are envisioning here. Would you mind stepping through an example (mapping [proposed] libvirt API calls to lvm commands would be helpful), on how an original lvm partition is created, then a snapshot partition, then how you run out of volume size in that snapshot? It sounds like you are saying that there is a way to create an lvm snapshot that is valid at the time that is created, but later on, subsequent actions cause the snapshot to run out of space and no longer be valid. But my understanding is that a snapshot is a constant size (it represents a known state of the disk at a fixed point in time), only the deltas to that snapshot (aka the live disk) ever have the potential to grow beyond the amount of storage used by the snapshot. Or are you worried about creating an lvm snapshot by libvirt, but then a third party program changes the property of the lvm snapshot volume to change its size?
Or is it something like the following scenario: start with 10G lvm partition create snapshot A to a 10GB partition make 2G worth of changes create snapshot B to a 2GB partition decide to delete snapshot A (merging it back into B) at that point, the 10G partition for the live file is still adequate, but now the effort of merging the 10G of data from snapshot into the 2G of storage reserved for snapshot B will fail, because the partition for B is not large enough. But I still don't see how that needs a new API; it seems like this would be a case of making the 'delete A' operation fail rather than a case of silently making snapshot B invalid. If the only way to make a snapshot lvm partition invalid is by making changes to that lvm partition outside the knowledge of libvirt, then I don't know that libvirt can protect itself against such 3rd party actions; and if libvirt's own actions can never cause an invalidation, then what is the point of adding an API to detect an invalid snapshot? But I'm relatively inexperienced with lvm, so an example of what you mean will go a long way to help me understand the scenario you are envisioning. -- Eric Blake eblake@redhat.com +1-801-349-2682 Libvirt virtualization library http://libvirt.org
Hi, 2011年6月21日7:52 Eric Blake <eblake@redhat.com>:
For LVM, it may be lvconvert --merge , but requires recent version of lvm and linux kernel, So, there are some failure reason (toolchain doesn't support it or
toolchain
supports, but failed) I don't know how to handle these case for now, but we may need to add some virErrorNumber to indicate snapshot-related error.
Yes, we can add new virError values as we come across situations where such operations are unsupported, but which don't fit well into any existing error values.
I understand.
virStorageVolSnapshotCreateXML. [And since my first phase of implementation will be focused on inline qcow2 snapshots, I don't yet know what that XML will need to contain for any other type of snapshots, such as mapping out how the snapshot backing file will be named in relation to the possibly new live file.]
In LVM case, taking snapshot(lvcreate) needs at least snapshot volume size as an argument, So, I think storage-specific element can be inserted in this XML format
Wouldn't the snapshot volume size always be equal to the volume size of the original? Or is there really value in allowing the snapshot size to be configured to something different (and if different, must it always be <= original, or does a larger snapshot than the original make sense)?
My concern senario is 1. creaing a snapshot through libvirt, 2. take a backup from snapshot, 3. remove a snapshot. case. In this case, I think sys-admin may want to create a snapshot size smaller than original size. But this senario seems to make another problem (snapshot becomes invalid, described below). so assuming that the snapshot volume size always be equal to the volume size of the original is better for us.
Any feedback on this approach? Any other APIs that would be useful to
add? I'd like to get all the new APIs in place for 0.9.3 with minimal qcow2 functionality, then use the time before 0.9.4 to further enhance the APIs to cover more snapshot cases but without having to add any new APIs.
LVM snapshot may become invalid in the case of running out the volume size, So, adding an API for checking whether snapshot is valid will be needed.
I'm not quite sure I follow the scenario you are envisioning here. Would you mind stepping through an example (mapping [proposed] libvirt API calls to lvm commands would be helpful), on how an original lvm partition is created, then a snapshot partition, then how you run out of volume size in that snapshot? It sounds like you are saying that there is a way to create an lvm snapshot that is valid at the time that is created, but later on, subsequent actions cause the snapshot to run out of space and no longer be valid. But my understanding is that a snapshot is a constant size (it represents a known state of the disk at a fixed point in time), only the deltas to that snapshot (aka the live disk) ever have the potential to grow beyond the amount of storage used by the snapshot. Or are you worried about creating an lvm snapshot by libvirt, but then a third party program changes the property of the lvm snapshot volume to change its size?
I worried about the case I mensioned above (snapshot size is smaller case) Regards,
On 06/15/2011 11:41 PM, Eric Blake wrote:
/* Opaque type to manage a snapshot of a single storage volume. */ typedef virStorageVolSnapshotPtr;
/* Create a snapshot of a storage volume. XML is optional, if non-NULL, it would be a new top-level element <volsnapshot> which is similar to the top-level <domainsnapshot> for virDomainSnapshotCreateXML, to specify name and description. Flags is 0 for now. */ virStorageVolSnapshotPtr virDomainSnapshotCreateXML(virStorageVolPtr vol, const char *xml, unsigned int flags);
Copy-and-paste error; should be: virStorageVolSnapshotPtr virStorageVolSnapshotCreateXML(virStorageVolPtr vol, const char *xml, unsigned int flags); -- Eric Blake eblake@redhat.com +1-801-349-2682 Libvirt virtualization library http://libvirt.org
On Wed, Jun 15, 2011 at 11:41:27PM -0600, Eric Blake wrote:
Right now, libvirt has a snapshot API via virDomainSnapshotCreateXML, but for qemu domains, it only works if all the guest disk images are qcow2, and qemu rather than libvirt does all the work. However, it has a couple of drawbacks: it is inherently tied to domains (there is no way to manage snapshots of storage volumes not tied to a domain, even though libvirt does that for qcow2 images associated with offline qemu domains by using the qemu-img application). And it necessarily operates on all of the images associated with a domain in parallel - if any disk image is not qcow2, the snapshot fails, and there is no way to select a subset of disks to save. However, it works on both active (disk and memory state) and inactive domains (just disk state).
Upstream qemu is developing a 'live snapshot' feature, which allows the creation of a snapshot without the current downtime of several seconds required by the current 'savevm' monitor command, as well as means for controlling applications (libvirt) to request that qemu pause I/O to a particular disk, then externally perform a snapshot, then tell qemu to resume I/O (perhaps on a different file name or fd from the host, but with no change to the contents seen by the guest). Eventually, these changes will make it possible for libvirt to create fast snapshots of LVM partitions or btrfs files for guest disk images, as well as to
Actually, IIUC, the QEMU 'live snapshot' feature is only for special disk formats like qcow2, qed, etc. For formats like LVM, brtfs, SCSI, etc, libvirt will have todo all the work of creating the snapshot, possibly then telling QEMU to switch the backing file of a virtual disk to the new image (if the snapshot mechanism works that way).
select which disks are saved in a snapshot (that is, save a crash-consistent state of a subset of disks, without the corresponding RAM state, rather than making a full system restore point); the latter would work best with guest cooperation to quiesce disks before qemu pauses I/O to that disk, but that is an orthogonal enhancement.
At the very least, you need a way to store QEMU writing to the disk for a period of time, whether or not the guest is quiesced. There are basically 3 options 1. Pause the guest CPUs (eg 'stop' on the monitor) 2. QEMU queues I/O from guest in memory temporarily (does not currently exist) 3. QEMU tells guest to quiesce I/O temporarily (does not currently exist) To perform a snapshot libvirt would need todo 1. Stop I/O using one of the 3 methods above 2. If disk is a special format - Ask QEMU to snapshot it Else - Create snapshot ourselves - Update QEMU disk backing path (optional) 3. Resume I/O
However, my first goal with API enhancements is to merely prove that libvirt can manage a live snapshot by using qemu-img on a qcow2 image rather than the current 'savevm' approach of qemu doing all the work.
FYI, QEMU developers are adament that if the disk image is open by QEMU you should, in general, not do anything using qemu-img on that disk image. libvirt does currently do things like querying disk capacity, but we can get away with that because it is an invariant section of the header. We certainly can't create internal snapshots with qemu-img while the guest is live. Creating external snapshots with qemu-img is probably OK, but when I've suggested this before QEMU developers were unhappy with even that.
Additionally, libvirt provides the virDomainSave command, which saves just the state of the domain's memory, and stops the guest. A crude libvirt-only snapshot could thus already be done by using virDomainSave, then externally doing a snapshot of all disk images associated with the domain by using virStorageVol APIs, except that such APIs don't yet exist. Additionally, virDomainSave has no flags argument, so there is no way to request that the guest be resumed after the snapshot completes.
Right now, I'm proposing the addition of virDomainSaveFlags, along with a series of virStorageVolSnapshot* APIs that mirror the virDomainSnapshot* APIs. This would mean adding:
/* Opaque type to manage a snapshot of a single storage volume. */ typedef virStorageVolSnapshotPtr;
/* Create a snapshot of a storage volume. XML is optional, if non-NULL, it would be a new top-level element <volsnapshot> which is similar to the top-level <domainsnapshot> for virDomainSnapshotCreateXML, to specify name and description. Flags is 0 for now. */ virStorageVolSnapshotPtr virDomainSnapshotCreateXML(virStorageVolPtr vol, const char *xml, unsigned int flags); [For qcow2, this would be implemented with 'qemu-img snapshot -c', similar to what virDomainSnapshotXML already does on inactive domains. Later, we can add LVM and btrfs support, or even allow full file copies of any file type. Also in the future, we could enhance XML to take a new element that describes a relationship between the name of the original and of the snapshot, in the case where a new filename has to be created to complete the snapshot process.]
/* Probe if vol has snapshots. 1 if true, 0 if false, -1 on error. Flags is 0 for now. */ int virStorageVolHasCurrentSnapshot(virStorageVolPtr vol, unsigned int flags); [For qcow2 images, snapshots can be contained within the same file and managed with qemu-img -l, but for other formats, this may mean that libvirt has to start managing externally saved data associated with the storage pool that associates snapshots with filenames. In fact, even for qcow2 it might be useful to support creation of new files backed by the previous snapshot rather than cramming multiple snapshots in one file, so we may have a use for flags to filter out the presence of single-file vs. multiple-file snapshot setups.]
/* Revert a volume back to the state of a snapshot, returning 0 on success. Flags is 0 for now. */ int virStorageVolRevertToSnapsot(virStorageVolSnapshotPtr snapshot, unsigned int flags); [For qcow2, this would involve qemu-img snapshot -a. Here, a useful flag might be whether to delete any changes made after the point of the snapshot; virDomainRevertToSnapshot should probably honor the same type of flag.]
/* Return the most recent snapshot of a volume, if one exists, or NULL on failure. Flags is 0 for now. */ virStorageVolSnapshotPtr virStorageVolSnapshotCurrent(virStorageVolPtr vol, unsigned int flags);
/* Delete the storage associated with a snapshot (although the opaque snapshot object must still be independently freed). If flags is 0, any child snapshots based off of this one are rebased onto the parent; if flags is VIR_STORAGE_VOL_SNAPSHOT_DELETE_CHILDREN , then any child snapshots based off of this one are also deleted. */ int virStorageVolSnapshotDelete(virStorageVolSnapshotPtr snapshot, unsigned int flags); [For qcow2, this would involve qemu-img snapshot -d. For multiple-file snapshots, this would also involve qemu-img commit.]
/* Free the object returned by virStorageVolSnapshot{Current,CreateXML,LookupByName}. The storage snapshot associated with this object still exists, if it has not been deleted by virStorageVolSnapshotDelete. */ int virStorageVolSnapshotFree(virStorageVolSnapshotPtr snapshot);
/* Return the <volsnapshot> XML details about this snapshot object. Flags is 0 for now. */ int virStorageVolSnapshotGetXMLDesc(virStorageVolSnapshotPtr snapshot, unsigned int flags);
/* Return the names of all snapshots associated with this volume, using len from virStorageVolSnapshotLen. Flags is 0 for now. */ int virStorageVolSnapshotListNames(virStorageVolPtr vol, char **names, int nameslen, unsigned int flags); [For qcow2, this involves qemu-img -l. Additionally, if virStorageVolHasCurrentSnapshot learns to filter on in-file vs. multi-file snapshots, then the same flags would apply here.]
/* Get the opaque object tied to a snapshot name. Flags is 0 for now. */ virStorageVolSnapshotPtr virStorageVolSnapshotLookupByName(virStorageVolPtr vol, const char *name, unsigned int flags);
/* Determine how many snapshots are tied to a volume, or -1 on error. Flags is 0 for now. */ int virStorageVolSnapshotNum(virStorageVolPtr vol, unsigned int flags); [Same flags as for virStorageVolSnapshotListNames.]
/* Save a domain into the file 'to' with additional actions. If flags is 0, then xml is ignored, and this is like virDomainSave. If flags includes VIR_DOMAIN_SAVE_DISKS, then all of the associated disk images are also snapshotted, as if by virStorageVolSnapshotCreateXML; the xml argument is optional, but if present, it should be a <domainsnapshot> element with <disk> sub-elements for directions on each disk that needs a non-empty xml argument for proper volume snapshot creation. If flags includes VIR_DOMAIN_SAVE_RESUME, then the guest is resumed after the offline snapshot is complete (note that VIR_DOMAIN_SAVE_RESUME without VIR_DOMAIN_SAVE_DISKS makes little sense, as a saved state file is rendered useless if the disk images are modified before it is resumed). If flags includes VIR_DOMAIN_SAVE_QUIESCE, this requests that a guest agent quiesce disk state before the saved state file is created. */ int virDomainSaveFlags(virDomainPtr domain, const char *to, const char *xml, unsigned int flags);
What I'm not seeing here, is how these APIs all relate to the existing support we have in virStorageVol APIs for creating snapshots. THis is already implemented for LVM, QCow, QCow2. The snapshots are created by specifying a backing file in the initial volume description. Depending on the storage type, the backing file for a snapshot can be writable, or readonly. Snapshots appear as just more storage volumes, and are not restricted to being within the same pool as the original volume. You can also mix storage formats, eg, create a Qcow2 volume with backing file on LVM, which is itself a snapshot of another LVM volume. The QCow2 internal snapshots don't really fit into our existing model, since they don't have extra associated external files, so maybe we do still want some of these explicit APIs to query snapshots against volumes.
Also, the existing virDomainSnapshotCreateXML can be made more powerful by adding new flags and enhancing the existing XML for <domainsnapshot>. When flags is 0, the current behavior of saving memory state alongside all disks (for running domains, via savevm) or just snapshotting all disks with default settings (for offline domains, via qemu-img) is kept. If flags includes VIR_DOMAIN_SNAPSHOT_LIVE, then the guest must be running, and the new monitor commands for live snapshots are used. If flags includes VIR_DOMAIN_SNAPSHOT_DISKS_ONLY, then only the disks are snapshotted (on a running guest, this generally means they will only be crash-consistent, and will need an fsck before that disk state can be remounted), but it will shave off time by not saving memory. If flags includes VIR_DOMAIN_SNAPSHOT_QUIESCE, then this will additionally request that a guest agent quiesce disk state before the live snapshot is taken (increasing the likelihood of a stable disk, rather than a crash-consistent disk; but it requires cooperation from the guest so it is no more reliable than memballoon changes).
As for the XML changes, it makes sense to snapshot just a subset of disks when you only care about crash-consistent state or if you can rely on a guest agent to quiesce the subset of disk(s) you care about, so the existing <domainsnapshot> element needs a new optional subelement to control which disks are snapshotted; additionally, this subelement will be useful for disk image formats that require additional complexity (such as a secondary file name, rather than the inline snapshot feature of qcow2). I'm envisioning something like the following:
<domainsnapshot> <name>whatever</name> <disk name='/path/to/image1' snapshot='no'/> <disk name='/path/to/image2'> <volsnapshot>...</volsnapshot> </disk> </domainsnapshot>
where there can be up to as many <disk> elements as there are disk <devices> in the domain xml; any domain disk not listed is given default treatment. The name attribute of <disk> is mandatory, in order to match this disk element to one of the domain disks. The snapshot='yes|no' attribute is optional, defaulting to yes, in order to skip a particular disk. The <volsnapshot> subelement is optional, but if present, it would be the same XML as is provided to the virStorageVolSnapshotCreateXML. [And since my first phase of implementation will be focused on inline qcow2 snapshots, I don't yet know what that XML will need to contain for any other type of snapshots, such as mapping out how the snapshot backing file will be named in relation to the possibly new live file.]
Any feedback on this approach? Any other APIs that would be useful to add? I'd like to get all the new APIs in place for 0.9.3 with minimal qcow2 functionality, then use the time before 0.9.4 to further enhance the APIs to cover more snapshot cases but without having to add any new APIs.
Daniel -- |: http://berrange.com -o- http://www.flickr.com/photos/dberrange/ :| |: http://libvirt.org -o- http://virt-manager.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :|
On Tue, Jun 21, 2011 at 11:30 AM, Daniel P. Berrange <berrange@redhat.com> wrote:
On Wed, Jun 15, 2011 at 11:41:27PM -0600, Eric Blake wrote:
Right now, libvirt has a snapshot API via virDomainSnapshotCreateXML, but for qemu domains, it only works if all the guest disk images are qcow2, and qemu rather than libvirt does all the work. However, it has a couple of drawbacks: it is inherently tied to domains (there is no way to manage snapshots of storage volumes not tied to a domain, even though libvirt does that for qcow2 images associated with offline qemu domains by using the qemu-img application). And it necessarily operates on all of the images associated with a domain in parallel - if any disk image is not qcow2, the snapshot fails, and there is no way to select a subset of disks to save. However, it works on both active (disk and memory state) and inactive domains (just disk state).
Upstream qemu is developing a 'live snapshot' feature, which allows the creation of a snapshot without the current downtime of several seconds required by the current 'savevm' monitor command, as well as means for controlling applications (libvirt) to request that qemu pause I/O to a particular disk, then externally perform a snapshot, then tell qemu to resume I/O (perhaps on a different file name or fd from the host, but with no change to the contents seen by the guest). Eventually, these changes will make it possible for libvirt to create fast snapshots of LVM partitions or btrfs files for guest disk images, as well as to
Actually, IIUC, the QEMU 'live snapshot' feature is only for special disk formats like qcow2, qed, etc.
Yes. The live snapshot feature in QEMU will not do btrfs, LVM, or SAN/NAS snapshots.
For formats like LVM, brtfs, SCSI, etc, libvirt will have todo all the work of creating the snapshot, possibly then telling QEMU to switch the backing file of a virtual disk to the new image (if the snapshot mechanism works that way).
Putting non-virtualization storage management code into libvirt seems suboptimal since other applications may also want to use these generic features. However, I'm not aware of a storage management API for Linux that would support LVM and various SAN/NAS appliances. Ideally we would have something like that and libvirt can use the storage management API without knowing all the different storage types. A service like udisks with plugins for SAN/NAS appliances could solve the problem of where to put the storage management code.
select which disks are saved in a snapshot (that is, save a crash-consistent state of a subset of disks, without the corresponding RAM state, rather than making a full system restore point); the latter would work best with guest cooperation to quiesce disks before qemu pauses I/O to that disk, but that is an orthogonal enhancement.
At the very least, you need a way to store QEMU writing to the disk for a period of time, whether or not the guest is quiesced. There are basically 3 options
1. Pause the guest CPUs (eg 'stop' on the monitor) 2. QEMU queues I/O from guest in memory temporarily (does not currently exist) 3. QEMU tells guest to quiesce I/O temporarily (does not currently exist)
To perform a snapshot libvirt would need todo
1. Stop I/O using one of the 3 methods above 2. If disk is a special format - Ask QEMU to snapshot it Else - Create snapshot ourselves - Update QEMU disk backing path (optional) 3. Resume I/O
Yes, QEMU needs to provide commands for these individual steps. Also, the guest must be notified of the snapshot operation so that it can flush in-memory data to disk - otherwise this cannot be used for backup purposes since guests with several GBs of RAM will keep a considerable portion of state in memory and disk will be out-of-date.
/* Save a domain into the file 'to' with additional actions. If flags is 0, then xml is ignored, and this is like virDomainSave. If flags includes VIR_DOMAIN_SAVE_DISKS, then all of the associated disk images are also snapshotted, as if by virStorageVolSnapshotCreateXML; the xml argument is optional, but if present, it should be a <domainsnapshot> element with <disk> sub-elements for directions on each disk that needs a non-empty xml argument for proper volume snapshot creation. If flags includes VIR_DOMAIN_SAVE_RESUME, then the guest is resumed after the offline snapshot is complete (note that VIR_DOMAIN_SAVE_RESUME without VIR_DOMAIN_SAVE_DISKS makes little sense, as a saved state file is rendered useless if the disk images are modified before it is resumed). If flags includes VIR_DOMAIN_SAVE_QUIESCE, this requests that a guest agent quiesce disk state before the saved state file is created. */ int virDomainSaveFlags(virDomainPtr domain, const char *to, const char *xml, unsigned int flags);
What I'm not seeing here, is how these APIs all relate to the existing support we have in virStorageVol APIs for creating snapshots. THis is already implemented for LVM, QCow, QCow2. The snapshots are created by specifying a backing file in the initial volume description. Depending on the storage type, the backing file for a snapshot can be writable, or readonly. Snapshots appear as just more storage volumes, and are not restricted to being within the same pool as the original volume. You can also mix storage formats, eg, create a Qcow2 volume with backing file on LVM, which is itself a snapshot of another LVM volume.
The QCow2 internal snapshots don't really fit into our existing model, since they don't have extra associated external files, so maybe we do still want some of these explicit APIs to query snapshots against volumes.
There will be restrictions based on the storage type you choose. For example, LVM volumes can only be backed off other LVM volumes AFAIK. It would be nice to expose snapshots as volumes and allow them to be accessed using virStorageVolDownload() (or an equivalent API). One snapshot use case is a backup solution that wants to integrate virtualization support. They would need to talk to libvirt to take regular snapshots and copy out the data before deleting the snapshot. In addition, they require a dirty block tracking API for incremental backups where they avoid copying out the entire disk contents by only copying the disk blocks that have changed. Implementing dirty block tracking for image formats in QEMU is possible and has been discussed a little already. btrfs also supports dirty block tracking between transaction IDs. Some SAN/NAS appliances may also expose this information. Any thoughts on a dirty block tracking API that produces an extent list of dirty blocks between a snapshot and another snapshot/volume? I think virStorageVolDownload() could be used to copy out only the dirty blocks. Stefan
On Tue, Jun 21, 2011 at 02:12:28PM +0100, Stefan Hajnoczi wrote:
On Tue, Jun 21, 2011 at 11:30 AM, Daniel P. Berrange <berrange@redhat.com> wrote:
For formats like LVM, brtfs, SCSI, etc, libvirt will have todo all the work of creating the snapshot, possibly then telling QEMU to switch the backing file of a virtual disk to the new image (if the snapshot mechanism works that way).
Putting non-virtualization storage management code into libvirt seems suboptimal since other applications may also want to use these generic features. However, I'm not aware of a storage management API for Linux that would support LVM and various SAN/NAS appliances. Ideally we would have something like that and libvirt can use the storage management API without knowing all the different storage types.
A service like udisks with plugins for SAN/NAS appliances could solve the problem of where to put the storage management code.
Well there is multiple answers to this "sub-optimality" 1/ we can't really wait, there are some parallel developments I'm following (from a distance) which may provide some of this the key point is trying to keep some compatibility in the objects and terms of the API to be able to reuse them 2/ if we develop some code in libvirt for this is could be separated as a library once it's mature enough IMHO the point is making sure the way we represent things maps easilly with how other specs or library may do it, then we should be able to reuse them (e.g. CDMI snapshots http://cdmi.sniacloud.com/CDMI_Spec/14-Snapshots/14-Snapshots.htm ) Daniel -- Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ daniel@veillard.com | Rpmfind RPM search engine http://rpmfind.net/ http://veillard.com/ | virtualization library http://libvirt.org/
On Wed, Jun 22, 2011 at 9:27 AM, Daniel Veillard <veillard@redhat.com> wrote:
On Tue, Jun 21, 2011 at 02:12:28PM +0100, Stefan Hajnoczi wrote:
On Tue, Jun 21, 2011 at 11:30 AM, Daniel P. Berrange <berrange@redhat.com> wrote:
For formats like LVM, brtfs, SCSI, etc, libvirt will have todo all the work of creating the snapshot, possibly then telling QEMU to switch the backing file of a virtual disk to the new image (if the snapshot mechanism works that way).
Putting non-virtualization storage management code into libvirt seems suboptimal since other applications may also want to use these generic features. However, I'm not aware of a storage management API for Linux that would support LVM and various SAN/NAS appliances. Ideally we would have something like that and libvirt can use the storage management API without knowing all the different storage types.
A service like udisks with plugins for SAN/NAS appliances could solve the problem of where to put the storage management code.
Well there is multiple answers to this "sub-optimality" 1/ we can't really wait, there are some parallel developments I'm following (from a distance) which may provide some of this the key point is trying to keep some compatibility in the objects and terms of the API to be able to reuse them 2/ if we develop some code in libvirt for this is could be separated as a library once it's mature enough
IMHO the point is making sure the way we represent things maps easilly with how other specs or library may do it, then we should be able to reuse them (e.g. CDMI snapshots http://cdmi.sniacloud.com/CDMI_Spec/14-Snapshots/14-Snapshots.htm )
I agree. The main thing is getting the semantics and the model right so that it can represent the various storage types/APIs. Stefan
On 06/21/2011 04:30 AM, Daniel P. Berrange wrote:
Upstream qemu is developing a 'live snapshot' feature, which allows the creation of a snapshot without the current downtime of several seconds required by the current 'savevm' monitor command, as well as means for controlling applications (libvirt) to request that qemu pause I/O to a particular disk, then externally perform a snapshot, then tell qemu to resume I/O (perhaps on a different file name or fd from the host, but with no change to the contents seen by the guest). Eventually, these changes will make it possible for libvirt to create fast snapshots of LVM partitions or btrfs files for guest disk images, as well as to
Actually, IIUC, the QEMU 'live snapshot' feature is only for special disk formats like qcow2, qed, etc.
Does anyone have pointers to the qemu implementation of monitor commands used for live snapshot?
For formats like LVM, brtfs, SCSI, etc, libvirt will have todo all the work of creating the snapshot, possibly then telling QEMU to switch the backing file of a virtual disk to the new image (if the snapshot mechanism works that way).
Yes, that was what I was envisioning.
select which disks are saved in a snapshot (that is, save a crash-consistent state of a subset of disks, without the corresponding RAM state, rather than making a full system restore point); the latter would work best with guest cooperation to quiesce disks before qemu pauses I/O to that disk, but that is an orthogonal enhancement.
At the very least, you need a way to store QEMU writing to the disk for a period of time, whether or not the guest is quiesced. There are basically 3 options
1. Pause the guest CPUs (eg 'stop' on the monitor) 2. QEMU queues I/O from guest in memory temporarily (does not currently exist) 3. QEMU tells guest to quiesce I/O temporarily (does not currently exist)
To perform a snapshot libvirt would need todo
1. Stop I/O using one of the 3 methods above 2. If disk is a special format - Ask QEMU to snapshot it Else - Create snapshot ourselves - Update QEMU disk backing path (optional) 3. Resume I/O
It is step 2B (create the snapshot ourselves) where the proposed virStorageVolSnapshot* APIs would be useful. The remaining steps also need implementation, but I believe that they can fit into existing APIs by the use of new flag values, rather than requiring any new API.
However, my first goal with API enhancements is to merely prove that libvirt can manage a live snapshot by using qemu-img on a qcow2 image rather than the current 'savevm' approach of qemu doing all the work.
FYI, QEMU developers are adament that if the disk image is open by QEMU you should, in general, not do anything using qemu-img on that disk image.
Agreed. And I further think that we need to expend some efforts making the new image locking code also play well with libvirt - that is, any virStorageVol API that can modify a disk image (rather than just do a read-only operation describing the image) should probably be taught to fail if any active domain is also using that image. Conversely, if a long-running virStorageVol API is started on a volume, then an attempt to virDomainStart a domain should see that the volume is already in use and fail just as if the volume had been locked by another running domain.
libvirt does currently do things like querying disk capacity, but we can get away with that because it is an invariant section of the header. We certainly can't create internal snapshots with qemu-img while the guest is live. Creating external snapshots with qemu-img is probably OK, but when I've suggested this before QEMU developers were unhappy with even that.
Basically, my proposed virStorageVolSnapshot APIs should only be used on inactive volumes; for a running domain, you should always go through the existing virDomainSnapshot API, which can then make appropriate decisions whether to do external snapshots, or whether to have qemu do the work because the image is qcow2. I think we're in agreement here, and that it still doesn't impact the decision for adding new API for offline snapshot management.
What I'm not seeing here, is how these APIs all relate to the existing support we have in virStorageVol APIs for creating snapshots. THis is already implemented for LVM, QCow, QCow2.
The only existing snapshot API that I found was virDomainSnapshotCreateXML, which only works on qcow2 (not lvm or qcow), and which works either online (via qemu) or offline (via qemu-img). But I could have overlooked something - where is the existing API for creating an LVM snapshot? For volume creation, I'm aware of code for specifying a backing file for an existing file, but backing files aren't necessarily the same as snapshots, are they?
The snapshots are created by specifying a backing file in the initial volume description. Depending on the storage type, the backing file for a snapshot can be writable, or readonly. Snapshots appear as just more storage volumes, and are not restricted to being within the same pool as the original volume. You can also mix storage formats, eg, create a Qcow2 volume with backing file on LVM, which is itself a snapshot of another LVM volume.
The QCow2 internal snapshots don't really fit into our existing model, since they don't have extra associated external files, so maybe we do still want some of these explicit APIs to query snapshots against volumes.
-- Eric Blake eblake@redhat.com +1-801-349-2682 Libvirt virtualization library http://libvirt.org
On Tue, Jun 21, 2011 at 2:53 PM, Eric Blake <eblake@redhat.com> wrote:
On 06/21/2011 04:30 AM, Daniel P. Berrange wrote:
Upstream qemu is developing a 'live snapshot' feature, which allows the creation of a snapshot without the current downtime of several seconds required by the current 'savevm' monitor command, as well as means for controlling applications (libvirt) to request that qemu pause I/O to a particular disk, then externally perform a snapshot, then tell qemu to resume I/O (perhaps on a different file name or fd from the host, but with no change to the contents seen by the guest). Eventually, these changes will make it possible for libvirt to create fast snapshots of LVM partitions or btrfs files for guest disk images, as well as to
Actually, IIUC, the QEMU 'live snapshot' feature is only for special disk formats like qcow2, qed, etc.
Does anyone have pointers to the qemu implementation of monitor commands used for live snapshot?
http://repo.or.cz/w/qemu.git/blob/HEAD:/blockdev.c#l572 Jes implemented the snapshot_blkdev command and is integrating guest agent fsfreeze support. I think it needs to be a multi-step process instead of just one command so that libvirt can do storage management for LVM and co. Stefan
On Thu, Jun 16, 2011 at 6:41 AM, Eric Blake <eblake@redhat.com> wrote: Robert, Fernando, Jagane: I have CCed you because we have discussed snapshot APIs and I thought you'd be interested in Eric's work to build them for libvirt. Does each volume have its own independent snapshot namespace? It may be wise to document that snapshot namespaces are *not* independent because storage backends may not be able to provide these semantics.
/* Create a snapshot of a storage volume. XML is optional, if non-NULL, it would be a new top-level element <volsnapshot> which is similar to the top-level <domainsnapshot> for virDomainSnapshotCreateXML, to specify name and description. Flags is 0 for now. */ virStorageVolSnapshotPtr virDomainSnapshotCreateXML(virStorageVolPtr vol, const char *xml, unsigned int flags); [For qcow2, this would be implemented with 'qemu-img snapshot -c', similar to what virDomainSnapshotXML already does on inactive domains. Later, we can add LVM and btrfs support, or even allow full file copies of any file type. Also in the future, we could enhance XML to take a new element that describes a relationship between the name of the original and of the snapshot, in the case where a new filename has to be created to complete the snapshot process.]
I found this formatting quite hard to read. Something along these lines would be easier to parse visually: /* Create a snapshot of a storage volume. XML is optional, if non-NULL, * it would be a new top-level element <volsnapshot> which is similar to * the top-level <domainsnapshot> for virDomainSnapshotCreateXML, to * specify name and description. Flags is 0 for now. */ virStorageVolSnapshotPtr virDomainSnapshotCreateXML( virStorageVolPtr vol, const char *xml, unsigned int flags); ...coments...
/* Probe if vol has snapshots. 1 if true, 0 if false, -1 on error. Flags is 0 for now. */ int virStorageVolHasCurrentSnapshot(virStorageVolPtr vol, unsigned int flags); [For qcow2 images, snapshots can be contained within the same file and managed with qemu-img -l, but for other formats, this may mean that libvirt has to start managing externally saved data associated with the storage pool that associates snapshots with filenames. In fact, even for qcow2 it might be useful to support creation of new files backed by the previous snapshot rather than cramming multiple snapshots in one file, so we may have a use for flags to filter out the presence of single-file vs. multiple-file snapshot setups.]
What is the "current snapshot"? Is this function necessary when you already have virStorageVolSnapshotListNames()?
/* Return the most recent snapshot of a volume, if one exists, or NULL on failure. Flags is 0 for now. */ virStorageVolSnapshotPtr virStorageVolSnapshotCurrent(virStorageVolPtr vol, unsigned int flags);
The name should include "revert". This looks like a shortcut function for virStorageVolRevertToSnapshot().
/* Delete the storage associated with a snapshot (although the opaque snapshot object must still be independently freed). If flags is 0, any child snapshots based off of this one are rebased onto the parent; if flags is VIR_STORAGE_VOL_SNAPSHOT_DELETE_CHILDREN , then any child snapshots based off of this one are also deleted. */
What is the "opaque snapshot object"?
int virStorageVolSnapshotDelete(virStorageVolSnapshotPtr snapshot, unsigned int flags); [For qcow2, this would involve qemu-img snapshot -d. For multiple-file snapshots, this would also involve qemu-img commit.]
Is virStorageVolDelete() modified to also delete the volume's snapshots? Stefan
Thanks for looping me in, Stefan.
Does each volume have its own independent snapshot namespace? It may be wise to document that snapshot namespaces are *not* independent because storage backends may not be able to provide these semantics.
There is a need for 'just-a-volume-snapshot', and for a 'whole-vm-snapshot'. The 'whole-vm-snapshot' can possibly be collection of 'just-a-volume-snapshot'.
/* Create a snapshot of a storage volume. XML is optional, if non-NULL, it would be a new top-level element<volsnapshot> which is similar to the top-level<domainsnapshot> for virDomainSnapshotCreateXML, to specify name and description. Flags is 0 for now. */ virStorageVolSnapshotPtr virDomainSnapshotCreateXML(virStorageVolPtr vol, const char *xml, unsigned int flags); [For qcow2, this would be implemented with 'qemu-img snapshot -c', similar to what virDomainSnapshotXML already does on inactive domains. Later, we can add LVM and btrfs support, or even allow full file copies of any file type. Also in the future, we could enhance XML to take a new element that describes a relationship between the name of the original and of the snapshot, in the case where a new filename has to be created to complete the snapshot process.] I found this formatting quite hard to read. Something along these lines would be easier to parse visually:
/* Create a snapshot of a storage volume. XML is optional, if non-NULL, * it would be a new top-level element<volsnapshot> which is similar to * the top-level<domainsnapshot> for virDomainSnapshotCreateXML, to * specify name and description. Flags is 0 for now. */ virStorageVolSnapshotPtr virDomainSnapshotCreateXML( virStorageVolPtr vol, const char *xml, unsigned int flags);
There are two types of snapshots that I am aware of: - Base file is left unmodified after snapshot, snapshot file is created and modified. e.g. qcow2 (I think) - Base file continues to be modified. The snapshot file gets COW blocks copied into it. e.g. LVM, Livebackup, etc. Can we enhance the libvirt API to indicate what type of snapshot is desired. Also, when a snapshot is listed, can we try and describe it as one kind or the other? There is no facility in the API to track dirty bitmaps. Suppose a disk format or qemu proper has the ability to maintain a dirty bitmap of blocks(or clusters) modified since some event (time in ms, perhaps). I would like libvirt to provide a function such as: /* * returns NULL if the underlying block driver does not support * maintaining a dirty bitmap. If it does support a dirty bitmap, * the driver returns an opaque object that represents the time * since which this dirty bitmap is valid. * * Used by incremental backup programs to determine if qemu * has a bitmap of blocks that were dirtied since the last time * a backup was taken. */ virStorageDirtyBitmapTimeOpaquePtr virStorageVolDirtyBitmapPresent(virStorageVolPtr vol) Thanks, Jagane
On Tue, Jul 5, 2011 at 2:53 AM, Jagane Sundar <jagane@sundar.org> wrote:
/* Create a snapshot of a storage volume. XML is optional, if non-NULL, * it would be a new top-level element<volsnapshot> which is similar to * the top-level<domainsnapshot> for virDomainSnapshotCreateXML, to * specify name and description. Flags is 0 for now. */ virStorageVolSnapshotPtr virDomainSnapshotCreateXML( virStorageVolPtr vol, const char *xml, unsigned int flags);
There are two types of snapshots that I am aware of: - Base file is left unmodified after snapshot, snapshot file is created and modified. e.g. qcow2 (I think)
More detail on this approach as implemented by QEMU's snapshot_blkdev: Create snapshot.qcow2 with base.img as backing file. base.img is now read-only and can be accessed as a "snapshot". All writes go to snapshot.qcow2. When the snapshot is no longer needed it is necessary to merge the COW data back into base.img before deleting snapshot.qcow2. This merge has not been implemented in QEMU yet.
- Base file continues to be modified. The snapshot file gets COW blocks copied into it. e.g. LVM, Livebackup, etc.
Can we enhance the libvirt API to indicate what type of snapshot is desired. Also, when a snapshot is listed, can we try and describe it as one kind or the other?
I think the snapshot mechanism will depend on your storage backend. If the disk image is an LVM volume, then it is natural for the snapshot to be an LVM snapshot. If the disk image is a qcow2 file, then it is natural for the snapshot be a QEMU snapshot_blkdev snapshot. Also, it is often not possible to mix these snapshot mechanisms. For example, LVM snapshots don't work on qcow2 image files. Does the application have to be aware of which snapshotting approach is used by the backend? Perhaps there are a few cases where it is technically possible to mix-and-match but it just seems to expose complexity without much gain. Put another way: "If a storage backend fundamentally doesn't support snapshotting the you like, use a different backend".
There is no facility in the API to track dirty bitmaps. Suppose a disk format or qemu proper has the ability to maintain a dirty bitmap of blocks(or clusters) modified since some event (time in ms, perhaps). I would like libvirt to provide a function such as:
/* * returns NULL if the underlying block driver does not support * maintaining a dirty bitmap. If it does support a dirty bitmap, * the driver returns an opaque object that represents the time * since which this dirty bitmap is valid. * * Used by incremental backup programs to determine if qemu * has a bitmap of blocks that were dirtied since the last time * a backup was taken. */ virStorageDirtyBitmapTimeOpaquePtr virStorageVolDirtyBitmapPresent(virStorageVolPtr vol)
Yes, dirty bitmap support is important. This will make backup much more efficient on storage backends that support it. For QEMU image files it will be possible to provide dirty block information in the future. btrfs and a SAN appliance that I have looked both have mechanisms that could be used to provide dirty block tracking. Stefan
On 07/05/2011 04:02 AM, Stefan Hajnoczi wrote:
On Tue, Jul 5, 2011 at 2:53 AM, Jagane Sundar <jagane@sundar.org> wrote:
/* Create a snapshot of a storage volume. XML is optional, if non-NULL, * it would be a new top-level element<volsnapshot> which is similar to * the top-level<domainsnapshot> for virDomainSnapshotCreateXML, to * specify name and description. Flags is 0 for now. */ virStorageVolSnapshotPtr virDomainSnapshotCreateXML( virStorageVolPtr vol, const char *xml, unsigned int flags);
There are two types of snapshots that I am aware of: - Base file is left unmodified after snapshot, snapshot file is created and modified. e.g. qcow2 (I think)
More detail on this approach as implemented by QEMU's snapshot_blkdev:
Create snapshot.qcow2 with base.img as backing file. base.img is now read-only and can be accessed as a "snapshot". All writes go to snapshot.qcow2.
When the snapshot is no longer needed it is necessary to merge the COW data back into base.img before deleting snapshot.qcow2.
or, to merge all of base.img into snapshot.qcow2 then change snapshot.qcow2 to no longer have a backing file, before deleting base.img. As I understand it, either file can be deleted when a snapshot is no longer needed, but having the flexibility to decide which of the two files to delete would be useful, and may require knowing how dirty the snapshot file is in relation to the original file (if it is 95% dirty, it is faster to just pull in the last few blocks from base into snapshot before deleting base, whereas if it is only 5% dirty, it is faster to sync the dirtied blocks from snapshot back to base before deleting snapshot). And even if you have control over which of the two images to delete, you may also want to have control over the final filename used for the merged image (that is, in the 5% dirty case, use the snapshot->base merge followed by rename(base,snapshot), rather than wasting time on the base->snapshot merge, to still get the end result that the final filename is snapshot).
This merge has not been implemented in QEMU yet.
Not to mention that it overlaps somewhat with the concept of live block copying.
- Base file continues to be modified. The snapshot file gets COW blocks copied into it. e.g. LVM, Livebackup, etc.
Can we enhance the libvirt API to indicate what type of snapshot is desired. Also, when a snapshot is listed, can we try and describe it as one kind or the other?
I think the snapshot mechanism will depend on your storage backend. If the disk image is an LVM volume, then it is natural for the snapshot to be an LVM snapshot. If the disk image is a qcow2 file, then it is natural for the snapshot be a QEMU snapshot_blkdev snapshot.
What if it is both at once? That is, it is possible to create an LVM partition whose contents are a qcow2 image. In that case, it seems like the user might want the flexibility to determine whether the snapshot is done at the qcow2 level or at the LVM level.
Also, it is often not possible to mix these snapshot mechanisms. For example, LVM snapshots don't work on qcow2 image files.
Why not? They might not be as space-efficient (the whole idea of LVM cloning is that each block of the original LVM partition is now COW-shared between multiple partitions, and that the backup partition only consumes as additional space according to the amount of blocks that get dirtied in the original partition), but I'm not seeing a technical reason that would prohibit them (and I welcome evidence to the contrary, so that I know more about what I am up against).
Does the application have to be aware of which snapshotting approach is used by the backend? Perhaps there are a few cases where it is technically possible to mix-and-match but it just seems to expose complexity without much gain.
Put another way: "If a storage backend fundamentally doesn't support snapshotting the you like, use a different backend".
So is this an accurate summary of your suggestion? vir{StorageVol,Domain}SnapshotGetXMLDesc should have a sub-element or attribute stating which method of snapshotting is in use, but other than telling you about the method, libvirt doesn't expose any further control over the matter (each disk gets snapshotted in the most efficient manner for that disk, given the constraints of the storage pool [directory vs. LVM partition] and image type [raw vs. qcow2 vs. qed] involved in that storage volume).
Yes, dirty bitmap support is important. This will make backup much more efficient on storage backends that support it.
For QEMU image files it will be possible to provide dirty block information in the future. btrfs and a SAN appliance that I have looked both have mechanisms that could be used to provide dirty block tracking.
-- Eric Blake eblake@redhat.com +1-801-349-2682 Libvirt virtualization library http://libvirt.org
The 05/07/11, Eric Blake wrote:
or, to merge all of base.img into snapshot.qcow2 then change snapshot.qcow2 to no longer have a backing file, before deleting base.img.
As I understand it, either file can be deleted when a snapshot is no longer needed, but having the flexibility to decide which of the two files to delete would be useful, and may require knowing how dirty the snapshot file is in relation to the original file (if it is 95% dirty, it is faster to just pull in the last few blocks from base into snapshot before deleting base, whereas if it is only 5% dirty, it is faster to sync the dirtied blocks from snapshot back to base before deleting snapshot). And even if you have control over which of the two images to delete, you may also want to have control over the final filename used for the merged image (that is, in the 5% dirty case, use the snapshot->base merge followed by rename(base,snapshot), rather than wasting time on the base->snapshot merge, to still get the end result that the final filename is snapshot).
I agree with your analysis. The current behaviour is blocking us from using qcow2 snapshots while your RFC fix the issues. But in your last sentence, what do you mean by "to still get the end result that the final filename is snapshot"? As end result, are you talking about: 1) the filename (which would mean we could have moving path for disks) or 2) the content (the snapshot content is in the final file, always in the path of the base file) ? -- Nicolas Sebrecht
On 07/06/2011 02:47 AM, Nicolas Sebrecht wrote:
And even if you have control over which of the two images to delete, you may also want to have control over the final filename used for the merged image (that is, in the 5% dirty case, use the snapshot->base merge followed by rename(base,snapshot), rather than wasting time on the base->snapshot merge, to still get the end result that the final filename is snapshot).
I agree with your analysis. The current behaviour is blocking us from using qcow2 snapshots while your RFC fix the issues.
But in your last sentence, what do you mean by "to still get the end result that the final filename is snapshot"? As end result, are you talking about:
1) the filename (which would mean we could have moving path for disks)
Starting from a single file base.img, if I take an external snapshot, I would now have base.img as the read-only base and new.img as the live file with a backing of base.img. If I then want to merge the two files, I want to choose which file to keep: if I want to keep base.img, then I either merge the dirty blocks from new.img back to base.img, or I merge the clean blocks from base.img to new.img then rename new.img to base.img. if I want to keep new.img, then I either merge clean blocks from base.img to new.img, or I merge dirty blocks from new.img to base.img then rename base.img to new.img. But your point about changing file names is also an important consideration - when libvirt tells qemu to do a snapshot, what is really happening (and should qemu support both modes of operation)? 1. libvirt pre-creates an empty file new.img with correct permissions, then tells qemu that file name; qemu uses the existing base.img as the read-only base and makes all further edits into the file new.img; so libvirt has to update the domain XML 2. libvirt creates a hard link new.img as an alternate name to base.img, then tells qemu that file name; qemu then opens new.img, unlinks base.img, and recreates base.img with new.img as the backing file, making all further edits to the new inode but existing base.img file name; so libvirt does not have to edit domain XML. Except that permissions may prevent qemu from re-creating a file, and truncating a hard-linked file is insufficient. So this method would involve some additional handshaking steps, where qemu would have to get help from libvirt in re-creating the new file. So this is a non-starter. 3. libvirt renames base.img to new.img while qemu still has the fd open, then creates base.img with the right permissions, and tells qemu to make the snapshot into base.img with a backing of new.img; all further edits go into the file base.img which now has a backing file of new.img. But this method implies that qemu has to either trust that libvirt did the rename correctly (or compare the inode between an fstat of the existing open fd and the stat of the backing file name), as well as implying that libvirt has to pass two filenames instead of one. It also implies that you can rename an in-use file (renaming devices doesn't work as well, and this isn't portable to mingw). So this also sounds like a non-starter. 4. Any other possibilities? In other words, it looks like we are stuck with updating XML to track new file names any time we take a snapshot. -- Eric Blake eblake@redhat.com +1-801-349-2682 Libvirt virtualization library http://libvirt.org
On Wed, Jul 6, 2011 at 3:03 PM, Eric Blake <eblake@redhat.com> wrote:
In other words, it looks like we are stuck with updating XML to track new file names any time we take a snapshot.
Yes, but QEMU's snapshot_blkdev command takes a filename argument so at least you get to specify that new filename. Stefan
On 07/07/2011 03:13 AM, Stefan Hajnoczi wrote:
On Wed, Jul 6, 2011 at 3:03 PM, Eric Blake <eblake@redhat.com> wrote:
In other words, it looks like we are stuck with updating XML to track new file names any time we take a snapshot.
Yes, but QEMU's snapshot_blkdev command takes a filename argument so at least you get to specify that new filename.
Well, the best thing (from libvirt's point of view) would be if snapshot_blkdev took a single string argument, which is either a /path/to/filename (and qemu does open()) or fd:name notation (to refer to a previously-named fd passed via the getfd monitor command, so that libvirt does open()). This would make SELinux integration easier, as one of the sVirt goals is to get to the point where we can use SELinux to forbid qemu from open()ing files on NFS shares, while still permitting all other operations on already-open fds passed in from libvirt. -- Eric Blake eblake@redhat.com +1-801-349-2682 Libvirt virtualization library http://libvirt.org
On Thu, Jul 7, 2011 at 8:34 PM, Eric Blake <eblake@redhat.com> wrote:
On 07/07/2011 03:13 AM, Stefan Hajnoczi wrote:
On Wed, Jul 6, 2011 at 3:03 PM, Eric Blake <eblake@redhat.com> wrote:
In other words, it looks like we are stuck with updating XML to track new file names any time we take a snapshot.
Yes, but QEMU's snapshot_blkdev command takes a filename argument so at least you get to specify that new filename.
Well, the best thing (from libvirt's point of view) would be if snapshot_blkdev took a single string argument, which is either a /path/to/filename (and qemu does open()) or fd:name notation (to refer to a previously-named fd passed via the getfd monitor command, so that libvirt does open()). This would make SELinux integration easier, as one of the sVirt goals is to get to the point where we can use SELinux to forbid qemu from open()ing files on NFS shares, while still permitting all other operations on already-open fds passed in from libvirt.
Today QEMU supports /path/to/filename. An fd argument to snapshot_blkdev requires a little bit of work since the QEMU block layer .bdrv_create() interface takes a filename and tries to create it. Jes: Is the fd argument to snapshot_blkdev in your plans? Stefan
On 07/08/11 10:58, Stefan Hajnoczi wrote:
On Thu, Jul 7, 2011 at 8:34 PM, Eric Blake <eblake@redhat.com> wrote:
Well, the best thing (from libvirt's point of view) would be if snapshot_blkdev took a single string argument, which is either a /path/to/filename (and qemu does open()) or fd:name notation (to refer to a previously-named fd passed via the getfd monitor command, so that libvirt does open()). This would make SELinux integration easier, as one of the sVirt goals is to get to the point where we can use SELinux to forbid qemu from open()ing files on NFS shares, while still permitting all other operations on already-open fds passed in from libvirt.
Today QEMU supports /path/to/filename. An fd argument to snapshot_blkdev requires a little bit of work since the QEMU block layer .bdrv_create() interface takes a filename and tries to create it.
Jes: Is the fd argument to snapshot_blkdev in your plans?
I only ever heard suggestions for taking fd arguments yesterday, so I cannot say it really is in my plans. If I get a good justification I might be convinced :) Cheers, Jes
On 07/08/2011 07:35 AM, Jes Sorensen wrote:
On 07/08/11 10:58, Stefan Hajnoczi wrote:
On Thu, Jul 7, 2011 at 8:34 PM, Eric Blake <eblake@redhat.com> wrote:
Well, the best thing (from libvirt's point of view) would be if snapshot_blkdev took a single string argument, which is either a /path/to/filename (and qemu does open()) or fd:name notation (to refer to a previously-named fd passed via the getfd monitor command, so that libvirt does open()). This would make SELinux integration easier, as one of the sVirt goals is to get to the point where we can use SELinux to forbid qemu from open()ing files on NFS shares, while still permitting all other operations on already-open fds passed in from libvirt.
Today QEMU supports /path/to/filename. An fd argument to snapshot_blkdev requires a little bit of work since the QEMU block layer .bdrv_create() interface takes a filename and tries to create it.
Jes: Is the fd argument to snapshot_blkdev in your plans?
I only ever heard suggestions for taking fd arguments yesterday, so I cannot say it really is in my plans. If I get a good justification I might be convinced :)
I already gave the justification - SELinux. For a disk mounted over NFS, the current SELinux bool virt_use_nfs is a bit broad (it gives all-or-nothing access to qemu to be able to open() arbitrary files on NFS). The way to tighten it is to instead have libvirt in charge of opening any file on NFS, then pass that fd to qemu - in which case SELinux policy can give qemu carte-blanche to do anything on an existing NFS fd, but not to open() a new one. Then sVirt has gained an additional layer of protection where qemu cannot blindly stomp on another VM's storage merely because both VMs happened to store their disk images on the same NFS server. But this requires that _all_ places in qemu that currently open() a file also be given an alternative to get the fd via command-line or getfd inheritance. -- Eric Blake eblake@redhat.com +1-801-349-2682 Libvirt virtualization library http://libvirt.org
On 07/04/2011 07:53 PM, Jagane Sundar wrote:
Thanks for looping me in, Stefan.
Does each volume have its own independent snapshot namespace? It may be wise to document that snapshot namespaces are *not* independent because storage backends may not be able to provide these semantics.
There is a need for 'just-a-volume-snapshot', and for a 'whole-vm-snapshot'. The 'whole-vm-snapshot' can possibly be collection of 'just-a-volume-snapshot'.
In the case of the current libvirt API, which makes use of the qemu 'savevm' monitor command (and thus is more of a checkpoint, rather than just disk snapshots): 'savevm' fails unless all disks associated with the guest are already qcow2 format. Additionally, it creates a snapshot visible to 'qemu-img snapshot -l' in all associated disks, but where only the primary disk additionally has the state of RAM also in the image. As for the creating a snapshot commands - the proposal for virStorageVolSnapshotCreateXML is _solely_ for offline management of storage volumes. If a storage volume is in use by a running qemu domain, then the only appropriate way to take an online snapshot of that disk (short of stopping the domain to get to the offline snapshot case) is to use the existing virDomainSnapshotCreateXML API instead. And that API is already flexible enough to support 'whole-vm-snapshot' vs. 'just-a-volume-snapshot'. The existing virDomainSnapshotCreateXML API is currently mapped to the 'savevm' command (which takes a checkpoint, which is the whole-vm-snapshot + memory), but can easily be modified to take just disk snapshots, and I already mentioned doing that by modifying the XML and adding a flag. That is, on creation: virDomainSnapshotCreateXML(domain, " <domainsnapshot> <name>whatever</name> </domainsnapshot> ", 0) is the existing usage, which creates a checkpoint (all qcow2 images get an internal snapshot named "whatever", and the first image also saves the memory state). virDomainSnapshotCreateXML(domain, " <domainsnapshot> <name>whatever</name> </domainsnapshot> ", VIR_DOMAIN_SNAPSHOT_DISK_ONLY) would try to create a snapshot of all disks associated with the image (although without the name of the snapshot file, either libvirt will have to have some default smarts for how to generate a reasonable backup file name, or this will fail). That is, omit all mention of <disk> subelements, and libvirt will then fill out the XML to cover all disks (the whole-vm-snapshot case). virDomainSnapshotCreateXML(domain, " <domainsnapshot> <name>whatever</name> <disk name='/path/to/image1' snapshot='no'/> <disk name='/path/to/image2'> <volsnapshot>...</volsnapshot> </disk> </domainsnapshot> ", VIR_DOMAIN_SNAPSHOT_DISK_ONLY) will only do a snapshot of disk image 2, using the <volsnapshot> information to explicitly specify the filename to use on the created external snapshot file (rather than letting libvirt generate the snapshot name). That is, provide at least one <disk> element, and you now have fine-grained control over which volumes get a snapshot (or even how that snapshot is created).
There are two types of snapshots that I am aware of: - Base file is left unmodified after snapshot, snapshot file is created and modified. e.g. qcow2 (I think) - Base file continues to be modified. The snapshot file gets COW blocks copied into it. e.g. LVM, Livebackup, etc.
There's a third - the qcow2 internal snapshot: - Base file contains both the snapshot and the delta.
Can we enhance the libvirt API to indicate what type of snapshot is desired. Also, when a snapshot is listed, can we try and describe it as one kind or the other?
Yes, there are already some read-only XML elements in the <domainsnapshot> XML (that is, libvirt ignores or rejects them if you pass them to virDomainSnapshotCreateXML, but virDomainSnapshotGetXMLDesc will list those additional elements to give you more details about the snapshot); having a sub-element to state whether the snapshot is backing-file based (original is now treated as read-only, and modifications affect the snapshot) or COW based (original and backup share all blocks to begin with, but as original get modified, the read-only backup has more unique blocks).
There is no facility in the API to track dirty bitmaps. Suppose a disk format or qemu proper has the ability to maintain a dirty bitmap of blocks(or clusters) modified since some event (time in ms, perhaps). I would like libvirt to provide a function such as:
/* * returns NULL if the underlying block driver does not support * maintaining a dirty bitmap. If it does support a dirty bitmap, * the driver returns an opaque object that represents the time * since which this dirty bitmap is valid. * * Used by incremental backup programs to determine if qemu * has a bitmap of blocks that were dirtied since the last time * a backup was taken. */ virStorageDirtyBitmapTimeOpaquePtr virStorageVolDirtyBitmapPresent(virStorageVolPtr vol)
Yes, we already had a discussion about the utility of being able to expose how much of an image is directly contained within a file, vs. being pulled in from a backing file (which can also be read as how much of an image is dirty compared to the state of a snapshot). See Daniel's earlier thoughts: https://www.redhat.com/archives/libvir-list/2011-April/msg00555.html -- Eric Blake eblake@redhat.com +1-801-349-2682 Libvirt virtualization library http://libvirt.org
On 7/5/2011 2:36 PM, Eric Blake wrote:
On 07/04/2011 07:53 PM, Jagane Sundar wrote:
Thanks for looping me in, Stefan.
Does each volume have its own independent snapshot namespace? It may be wise to document that snapshot namespaces are *not* independent because storage backends may not be able to provide these semantics.
There is a need for 'just-a-volume-snapshot', and for a 'whole-vm-snapshot'. The 'whole-vm-snapshot' can possibly be collection of 'just-a-volume-snapshot'. In the case of the current libvirt API, which makes use of the qemu 'savevm' monitor command (and thus is more of a checkpoint, rather than just disk snapshots):
'savevm' fails unless all disks associated with the guest are already qcow2 format. Additionally, it creates a snapshot visible to 'qemu-img snapshot -l' in all associated disks, but where only the primary disk additionally has the state of RAM also in the image.
As for the creating a snapshot commands - the proposal for virStorageVolSnapshotCreateXML is _solely_ for offline management of storage volumes. If a storage volume is in use by a running qemu domain, then the only appropriate way to take an online snapshot of that disk (short of stopping the domain to get to the offline snapshot case) is to use the existing virDomainSnapshotCreateXML API instead. And that API is already flexible enough to support 'whole-vm-snapshot' vs. 'just-a-volume-snapshot'.
OK. My use case is solved by virDomainSnapshotCreateXML, so long as this function is enhanced to support disk-only snapshots.
information to explicitly specify the filename to use on the created external snapshot file (rather than letting libvirt generate the snapshot name). That is, provide at least one<disk> element, and you now have fine-grained control over which volumes get a snapshot (or even how that snapshot is created).
There are two types of snapshots that I am aware of: - Base file is left unmodified after snapshot, snapshot file is created and modified. e.g. qcow2 (I think) - Base file continues to be modified. The snapshot file gets COW blocks copied into it. e.g. LVM, Livebackup, etc. There's a third - the qcow2 internal snapshot:
- Base file contains both the snapshot and the delta.
Can we enhance the libvirt API to indicate what type of snapshot is desired. Also, when a snapshot is listed, can we try and describe it as one kind or the other? Yes, there are already some read-only XML elements in the <domainsnapshot> XML (that is, libvirt ignores or rejects them if you pass them to virDomainSnapshotCreateXML, but virDomainSnapshotGetXMLDesc will list those additional elements to give you more details about the snapshot); having a sub-element to state whether the snapshot is backing-file based (original is now treated as read-only, and modifications affect the snapshot) or COW based (original and backup share all blocks to begin with, but as original get modified, the read-only backup has more unique blocks).
Sounds fine.
There is no facility in the API to track dirty bitmaps. Suppose a disk format or qemu proper has the ability to maintain a dirty bitmap of blocks(or clusters) modified since some event (time in ms, perhaps). I would like libvirt to provide a function such as:
/* * returns NULL if the underlying block driver does not support * maintaining a dirty bitmap. If it does support a dirty bitmap, * the driver returns an opaque object that represents the time * since which this dirty bitmap is valid. * * Used by incremental backup programs to determine if qemu * has a bitmap of blocks that were dirtied since the last time * a backup was taken. */ virStorageDirtyBitmapTimeOpaquePtr virStorageVolDirtyBitmapPresent(virStorageVolPtr vol) Yes, we already had a discussion about the utility of being able to expose how much of an image is directly contained within a file, vs. being pulled in from a backing file (which can also be read as how much of an image is dirty compared to the state of a snapshot). See Daniel's earlier thoughts: https://www.redhat.com/archives/libvir-list/2011-April/msg00555.html
You are right. The thread that you point to is indeed more relevant to my use case. I missed that thread of conversation. However, I do not see a conclusion to the discussion. My position is similar to Daniel's - I would find an API that exposes 'logical allocation information' (to quote Daniel) useful. His use case is for COR image streaming. My use case is for making incremental backups. The backup program wants to know which blocks were modified since the last backup was taken. Thanks, Jagane
On 07/04/2011 08:19 AM, Stefan Hajnoczi wrote:
On Thu, Jun 16, 2011 at 6:41 AM, Eric Blake <eblake@redhat.com> wrote:
Robert, Fernando, Jagane: I have CCed you because we have discussed snapshot APIs and I thought you'd be interested in Eric's work to build them for libvirt.
Does each volume have its own independent snapshot namespace? It may be wise to document that snapshot namespaces are *not* independent because storage backends may not be able to provide these semantics.
Good question, and I'm not quite sure on the best way to represent this. For qcow2 internal snapshots, the answer is obvious - each qcow2 image has its own snapshot namespace (and you can currently view this with qemu-img snapshot -l). But it looks like the existing drive to add qemu snapshot support to live domains is focusing solely on external snapshots (that is, a qcow2 snapshot involves creating a new filename, then marking the old filename as a read-only backing image of the new qcow2 filename; where qcow2 can even be used as a snapshot around a raw file). For external snapshots, I'm not sure whether the namespace should be specific to a storage pool (that is, any additional metadata that libvirt needs to track snapshot relationships should be stored in the same directory as the disk images themselves) or specific to the libvirt host (that is, just as libvirt's notion of a persistent storage pool happens to be stored in /etc/libvirt/storage/pool.xml, libvirt should also manage a file /etc/libvirt/storage/pool/snapshot.xml to describe all snapshots tracked within "pool"). But I'm certainly thinking that it is more likely to be a pool-wide namespace, rather than a storage-volume local namespace.
I found this formatting quite hard to read. Something along these lines would be easier to parse visually:
Yeah, sometimes there's only so much you can do with plain text emails. I'll try to make future revisions of this RFC easier to parse, and/or get it all in the wiki with nicer formatting first.
/* Probe if vol has snapshots. 1 if true, 0 if false, -1 on error. Flags is 0 for now. */ int virStorageVolHasCurrentSnapshot(virStorageVolPtr vol, unsigned int flags); [For qcow2 images, snapshots can be contained within the same file and managed with qemu-img -l, but for other formats, this may mean that libvirt has to start managing externally saved data associated with the storage pool that associates snapshots with filenames. In fact, even for qcow2 it might be useful to support creation of new files backed by the previous snapshot rather than cramming multiple snapshots in one file, so we may have a use for flags to filter out the presence of single-file vs. multiple-file snapshot setups.]
What is the "current snapshot"?
I was trying to model this after the virDomainSnapshot API, which has a notion of current snapshot, but I guess I don't fully understand what that API was implying by current domain snapshot. After looking more through the code, it looks like the idea was to support the notion of a branching hierarchy of snapshots: A -> B -> D \-> C -> E such that you can revert to the snapshot at point B, start the domain, and create snapshot D; then stop the domain, revert to the snapshot at point C, start the domain, and create snapshot E. Every time you start the domain, you know from which state it was started, so the current snapshot is the snapshot that would be the parent if you were to create a snapshot right now. As long as you can create a branching hierarchy (that is, after creating a snapshot, you can revert to its parent state, then make different changes, and create a new snapshot), then there really is a need to know which snapshot is current - listing all snapshots can show you the parent(s) of each snapshot in that list, but can't tell you which snapshot would be the parent of a new one created right now.
Is this function necessary when you already have virStorageVolSnapshotListNames()?
Maybe I need to understand why we have it for the virDomainSnapshot case, and whether it still makes sense for a disk image that is not associated with a domain. To some degree, I think it seems necessary, to reflect the fact that with internal qcow2 snapshots, I can do: qemu-img snapshsot -c one file run then stop vm to modify file qemu-img snapshot -c two file qemu-img snapshot -a one file run then stop vm to modify file qemu-img snapshot -c three file with the resulting hierarchy: one -> two \-> three On the other hand, qemu-img doesn't appear to list any hierarchies between internal snapshots - that is, while 'qemu-img snapshot -l' will list one, two, and three, it gives no indication that three depends on one but not two, nor whether the current state of the file would be a delta against three, two, one, or even parent-less. This also starts to get into questions about the ability to split a qcow2 image with internal snapshots. That is, if I have a single file with snapshot one and a delta against that snapshot as the current disk state, it would be nice to create a new qcow2 file with identical contents to snapshot one, then rebase the existing qcow2 file to have a backing file of my new clone file and delete the internal snapshot from the original file. But this starts to sound like work on live block copy APIs. For an offline storage volume, we can do things manually (qemu-img snapshot -c to temporarily create yet another snapshot point to later return to, qemu-img snapshot -a to revert to the snapshot of interest, then qemu-img convert to copy off the contents, then qemu-img snapshot -a to the temporary state, then qemu-img snapshot -d to clean up both the temporary andstate). But for a storage volume currently in use by qemu, this would imply a new qemu command to have qemu assist in streaming out the contents of the snapshot state.
/* Return the most recent snapshot of a volume, if one exists, or NULL on failure. Flags is 0 for now. */ virStorageVolSnapshotPtr virStorageVolSnapshotCurrent(virStorageVolPtr vol, unsigned int flags);
The name should include "revert". This looks like a shortcut function for virStorageVolRevertToSnapshot().
No, it was intended as a counterpart to virDomainSnapshotCurrent, which returns the "current snapshot" if there is one. But again, it may be that while a "current snapshot" makes sense for a domain, it might not make sense for a storage volume in isolation.
/* Delete the storage associated with a snapshot (although the opaque snapshot object must still be independently freed). If flags is 0, any child snapshots based off of this one are rebased onto the parent; if flags is VIR_STORAGE_VOL_SNAPSHOT_DELETE_CHILDREN , then any child snapshots based off of this one are also deleted. */
What is the "opaque snapshot object"?
Any instance of virDomainStorageSnapshotPtr. It is opaque because the caller does not need to know the contents of struct _virDomainStorageSnapshot.
int virStorageVolSnapshotDelete(virStorageVolSnapshotPtr snapshot, unsigned int flags); [For qcow2, this would involve qemu-img snapshot -d. For multiple-file snapshots, this would also involve qemu-img commit.]
Is virStorageVolDelete() modified to also delete the volume's snapshots?
I think this has come up elsewhere in the thread, but yes, virStorageVolDelete needs to be taught whether to error out if snapshots exist, or whether to also delete all snapshots associated with a storage volume being deleted. -- Eric Blake eblake@redhat.com +1-801-349-2682 Libvirt virtualization library http://libvirt.org
The 05/07/11, Eric Blake wrote:
I was trying to model this after the virDomainSnapshot API, which has a notion of current snapshot, but I guess I don't fully understand what that API was implying by current domain snapshot.
After looking more through the code, it looks like the idea was to support the notion of a branching hierarchy of snapshots:
A -> B -> D \-> C -> E
such that you can revert to the snapshot at point B, start the domain, and create snapshot D; then stop the domain, revert to the snapshot at point C, start the domain, and create snapshot E. Every time you start the domain, you know from which state it was started, so the current snapshot is the snapshot that would be the parent if you were to create a snapshot right now.
As long as you can create a branching hierarchy (that is, after creating a snapshot, you can revert to its parent state, then make different changes, and create a new snapshot), then there really is a need to know which snapshot is current - listing all snapshots can show you the parent(s) of each snapshot in that list, but can't tell you which snapshot would be the parent of a new one created right now.
May I ask what would happen if the user asks for deletion of A (as both base and snapshot)? Would it be possible? Would it be merged to both B and C? Using which filename? -- Nicolas Sebrecht
On 07/06/2011 02:59 AM, Nicolas Sebrecht wrote:
The 05/07/11, Eric Blake wrote:
I was trying to model this after the virDomainSnapshot API, which has a notion of current snapshot, but I guess I don't fully understand what that API was implying by current domain snapshot.
After looking more through the code, it looks like the idea was to support the notion of a branching hierarchy of snapshots:
A -> B -> D \-> C -> E
Someone pointed out to me (I'm not sure if it was in this thread) that qcow2 does _not_ by itself maintain any hierarchy - from qemu-img's point of view, all snapshots are equally independent reference-counted copies of data at a moment in time. So the hierarchy present in libvirt's use of virDomainSnapshot* APIs as applied to system checkpoints is a factor of libvirt maintaining extra metadata alongside the qcow2 images. That is, libvirt remembers which checkpoint was most recently created/reverted, and when creating a new checkpoint, assigns the current snapshot as the parent of the new one.
May I ask what would happen if the user asks for deletion of A (as both base and snapshot)? Would it be possible? Would it be merged to both B and C? Using which filename?
Libvirt already handles this just fine for live checkpoints (although in my testing yesterday, it look like libvirt forgets to maintain a notion of current snapshot if you used virDomainSnapshotCreateXML while a domain was inactive, making for an incomplete hierarchy). Basically, when deleting a snapshot, you have two options - delete that snapshot and all descendants, or delete just the snapshot and reparent all immediate children. That is, in the above scenario, when deleting B, you can choose whether to end up with: A -> C -> E (B and B's children all deleted) or A -> D \-> C -> E (B deleted, and D reparented to B's parent) Right now, libvirt's handling of checkpoint hierarchy under virDomainSnapshot* API fits in with the qemu 'savevm' usage, which uses only internal qcow2 snapshots. So there are no filenames to merge. You are indeed correct, though, that when deleting snapshots where snapshots involve external files, that it would be nice to have user control over which of the two files gets deleted. But the same principles apply - if E.img has C.img as a backing file, then requesting to delete snapshot C means that either E.img will also need to be deleted, or it will need to be rebased to have its backing-file updated to match the parent of C.img. -- Eric Blake eblake@redhat.com +1-801-349-2682 Libvirt virtualization library http://libvirt.org
On Tue, Jul 5, 2011 at 8:59 PM, Eric Blake <eblake@redhat.com> wrote:
On 07/04/2011 08:19 AM, Stefan Hajnoczi wrote:
On Thu, Jun 16, 2011 at 6:41 AM, Eric Blake <eblake@redhat.com> wrote:
Robert, Fernando, Jagane: I have CCed you because we have discussed snapshot APIs and I thought you'd be interested in Eric's work to build them for libvirt.
Does each volume have its own independent snapshot namespace? It may be wise to document that snapshot namespaces are *not* independent because storage backends may not be able to provide these semantics.
Good question, and I'm not quite sure on the best way to represent this.
For qcow2 internal snapshots, the answer is obvious - each qcow2 image has its own snapshot namespace (and you can currently view this with qemu-img snapshot -l). But it looks like the existing drive to add qemu snapshot support to live domains is focusing solely on external snapshots (that is, a qcow2 snapshot involves creating a new filename, then marking the old filename as a read-only backing image of the new qcow2 filename; where qcow2 can even be used as a snapshot around a raw file).
For external snapshots, I'm not sure whether the namespace should be specific to a storage pool (that is, any additional metadata that libvirt needs to track snapshot relationships should be stored in the same directory as the disk images themselves) or specific to the libvirt host (that is, just as libvirt's notion of a persistent storage pool happens to be stored in /etc/libvirt/storage/pool.xml, libvirt should also manage a file /etc/libvirt/storage/pool/snapshot.xml to describe all snapshots tracked within "pool"). But I'm certainly thinking that it is more likely to be a pool-wide namespace, rather than a storage-volume local namespace.
Pool-wide seems reasonable.
Is this function necessary when you already have virStorageVolSnapshotListNames()?
Maybe I need to understand why we have it for the virDomainSnapshot case, and whether it still makes sense for a disk image that is not associated with a domain. To some degree, I think it seems necessary, to reflect the fact that with internal qcow2 snapshots, I can do:
qemu-img snapshsot -c one file run then stop vm to modify file qemu-img snapshot -c two file qemu-img snapshot -a one file run then stop vm to modify file qemu-img snapshot -c three file
with the resulting hierarchy:
one -> two \-> three
On the other hand, qemu-img doesn't appear to list any hierarchies between internal snapshots - that is, while 'qemu-img snapshot -l' will list one, two, and three, it gives no indication that three depends on one but not two, nor whether the current state of the file would be a delta against three, two, one, or even parent-less.
There is no explicit relationship between internal qcow2 snapshots. qcow2 does reference counting of the actual data clusters and tables but the snapshot itself is oblivious. You can delete "one" without affecting "two" or "three". There is no dependency relationship between snapshots themselves, only reference counts on data clusters and tables. Here is how qcow2 snapshot operations work: 1. Create snapshot Increment reference counts for entire active image. Copy active L1 table into the snapshot data structure. 2. Activate snapshot Decrement reference counts for entire active image. Copy snapshot L1 table into active data structure. Increment reference counts for entire active image. 3. Delete snapshot Decrement reference counts for entire snapshot image.
This also starts to get into questions about the ability to split a qcow2 image with internal snapshots. That is, if I have a single file with snapshot one and a delta against that snapshot as the current disk state, it would be nice to create a new qcow2 file with identical contents to snapshot one, then rebase the existing qcow2 file to have a backing file of my new clone file and delete the internal snapshot from the original file. But this starts to sound like work on live block copy APIs. For an offline storage volume, we can do things manually (qemu-img snapshot -c to temporarily create yet another snapshot point to later return to, qemu-img snapshot -a to revert to the snapshot of interest, then qemu-img convert to copy off the contents, then qemu-img snapshot -a to the temporary state, then qemu-img snapshot -d to clean up both the temporary andstate). But for a storage volume currently in use by qemu, this would imply a new qemu command to have qemu assist in streaming out the contents of the snapshot state.
The current live block copy/image streaming APIs do not know about internal snapshots. Copy the contents of a snapshot while the VM is running is technically doable but there is no API and no code for it in QEMU.
/* Return the most recent snapshot of a volume, if one exists, or NULL on failure. Flags is 0 for now. */ virStorageVolSnapshotPtr virStorageVolSnapshotCurrent(virStorageVolPtr vol, unsigned int flags);
The name should include "revert". This looks like a shortcut function for virStorageVolRevertToSnapshot().
No, it was intended as a counterpart to virDomainSnapshotCurrent, which returns the "current snapshot" if there is one. But again, it may be that while a "current snapshot" makes sense for a domain, it might not make sense for a storage volume in isolation.
qcow2 internal snapshots always copy metadata to the "active" image, they do not allow you to in-place update existing snapshots. In that sense a qcow2 image only has one current snapshot, the active image. The way to update a snapshot is to delete it and create a new one with the same name. Stefan
On 06/16/2011 01:41 PM, Eric Blake wrote:
Right now, libvirt has a snapshot API via virDomainSnapshotCreateXML, but for qemu domains, it only works if all the guest disk images are qcow2, and qemu rather than libvirt does all the work. However, it has a couple of drawbacks: it is inherently tied to domains (there is no way to manage snapshots of storage volumes not tied to a domain, even though libvirt does that for qcow2 images associated with offline qemu domains by using the qemu-img application). And it necessarily operates on all of the images associated with a domain in parallel - if any disk image is not qcow2, the snapshot fails, and there is no way to select a subset of disks to save. However, it works on both active (disk and memory state) and inactive domains (just disk state).
Upstream qemu is developing a 'live snapshot' feature, which allows the creation of a snapshot without the current downtime of several seconds required by the current 'savevm' monitor command, as well as means for controlling applications (libvirt) to request that qemu pause I/O to a particular disk, then externally perform a snapshot, then tell qemu to resume I/O (perhaps on a different file name or fd from the host, but with no change to the contents seen by the guest). Eventually, these changes will make it possible for libvirt to create fast snapshots of LVM partitions or btrfs files for guest disk images, as well as to select which disks are saved in a snapshot (that is, save a crash-consistent state of a subset of disks, without the corresponding RAM state, rather than making a full system restore point); the latter would work best with guest cooperation to quiesce disks before qemu pauses I/O to that disk, but that is an orthogonal enhancement. However, my first goal with API enhancements is to merely prove that libvirt can manage a live snapshot by using qemu-img on a qcow2 image rather than the current 'savevm' approach of qemu doing all the work.
Additionally, libvirt provides the virDomainSave command, which saves just the state of the domain's memory, and stops the guest. A crude libvirt-only snapshot could thus already be done by using virDomainSave, then externally doing a snapshot of all disk images associated with the domain by using virStorageVol APIs, except that such APIs don't yet exist. Additionally, virDomainSave has no flags argument, so there is no way to request that the guest be resumed after the snapshot completes.
Right now, I'm proposing the addition of virDomainSaveFlags, along with a series of virStorageVolSnapshot* APIs that mirror the virDomainSnapshot* APIs. This would mean adding:
/* Opaque type to manage a snapshot of a single storage volume. */ typedef virStorageVolSnapshotPtr;
/* Create a snapshot of a storage volume. XML is optional, if non-NULL, it would be a new top-level element<volsnapshot> which is similar to the top-level<domainsnapshot> for virDomainSnapshotCreateXML, to specify name and description. Flags is 0 for now. */ virStorageVolSnapshotPtr virDomainSnapshotCreateXML(virStorageVolPtr vol, const char *xml, unsigned int flags); [For qcow2, this would be implemented with 'qemu-img snapshot -c', similar to what virDomainSnapshotXML already does on inactive domains. Later, we can add LVM and btrfs support, or even allow full file copies of any file type. Also in the future, we could enhance XML to take a new element that describes a relationship between the name of the original and of the snapshot, in the case where a new filename has to be created to complete the snapshot process.]
/* Probe if vol has snapshots. 1 if true, 0 if false, -1 on error. Flags is 0 for now. */ int virStorageVolHasCurrentSnapshot(virStorageVolPtr vol, unsigned int flags); [For qcow2 images, snapshots can be contained within the same file and managed with qemu-img -l, but for other formats, this may mean that libvirt has to start managing externally saved data associated with the storage pool that associates snapshots with filenames. In fact, even for qcow2 it might be useful to support creation of new files backed by the previous snapshot rather than cramming multiple snapshots in one file, so we may have a use for flags to filter out the presence of single-file vs. multiple-file snapshot setups.]
/* Revert a volume back to the state of a snapshot, returning 0 on success. Flags is 0 for now. */ int virStorageVolRevertToSnapsot(virStorageVolSnapshotPtr snapshot, unsigned int flags); [For qcow2, this would involve qemu-img snapshot -a. Here, a useful flag might be whether to delete any changes made after the point of the snapshot; virDomainRevertToSnapshot should probably honor the same type of flag.]
/* Return the most recent snapshot of a volume, if one exists, or NULL on failure. Flags is 0 for now. */ virStorageVolSnapshotPtr virStorageVolSnapshotCurrent(virStorageVolPtr vol, unsigned int flags);
/* Delete the storage associated with a snapshot (although the opaque snapshot object must still be independently freed). If flags is 0, any child snapshots based off of this one are rebased onto the parent; if flags is VIR_STORAGE_VOL_SNAPSHOT_DELETE_CHILDREN , then any child snapshots based off of this one are also deleted. */ int virStorageVolSnapshotDelete(virStorageVolSnapshotPtr snapshot, unsigned int flags); [For qcow2, this would involve qemu-img snapshot -d. For multiple-file snapshots, this would also involve qemu-img commit.]
/* Free the object returned by virStorageVolSnapshot{Current,CreateXML,LookupByName}. The storage snapshot associated with this object still exists, if it has not been deleted by virStorageVolSnapshotDelete. */ int virStorageVolSnapshotFree(virStorageVolSnapshotPtr snapshot);
/* Return the<volsnapshot> XML details about this snapshot object. Flags is 0 for now. */ int virStorageVolSnapshotGetXMLDesc(virStorageVolSnapshotPtr snapshot, unsigned int flags);
/* Return the names of all snapshots associated with this volume, using len from virStorageVolSnapshotLen. Flags is 0 for now. */ int virStorageVolSnapshotListNames(virStorageVolPtr vol, char **names, int nameslen, unsigned int flags); [For qcow2, this involves qemu-img -l. Additionally, if virStorageVolHasCurrentSnapshot learns to filter on in-file vs. multi-file snapshots, then the same flags would apply here.]
/* Get the opaque object tied to a snapshot name. Flags is 0 for now. */ virStorageVolSnapshotPtr virStorageVolSnapshotLookupByName(virStorageVolPtr vol, const char *name, unsigned int flags);
/* Determine how many snapshots are tied to a volume, or -1 on error. Flags is 0 for now. */ int virStorageVolSnapshotNum(virStorageVolPtr vol, unsigned int flags); [Same flags as for virStorageVolSnapshotListNames.]
/* Save a domain into the file 'to' with additional actions. If flags is 0, then xml is ignored, and this is like virDomainSave. If flags includes VIR_DOMAIN_SAVE_DISKS, then all of the associated disk images are also snapshotted, as if by virStorageVolSnapshotCreateXML; the xml argument is optional, but if present, it should be a<domainsnapshot> element with<disk> sub-elements for directions on each disk that needs a non-empty xml argument for proper volume snapshot creation. If flags includes VIR_DOMAIN_SAVE_RESUME, then the guest is resumed after the offline snapshot is complete (note that VIR_DOMAIN_SAVE_RESUME without VIR_DOMAIN_SAVE_DISKS makes little sense, as a saved state file is rendered useless if the disk images are modified before it is resumed). If flags includes VIR_DOMAIN_SAVE_QUIESCE, this requests that a guest agent quiesce disk state before the saved state file is created. */ int virDomainSaveFlags(virDomainPtr domain, const char *to, const char *xml, unsigned int flags);
Also, the existing virDomainSnapshotCreateXML can be made more powerful by adding new flags and enhancing the existing XML for<domainsnapshot>. When flags is 0, the current behavior of saving memory state alongside all disks (for running domains, via savevm) or just snapshotting all disks with default settings (for offline domains, via qemu-img) is kept. If flags includes VIR_DOMAIN_SNAPSHOT_LIVE, then the guest must be running, and the new monitor commands for live snapshots are used. If flags includes VIR_DOMAIN_SNAPSHOT_DISKS_ONLY, then only the disks are snapshotted (on a running guest, this generally means they will only be crash-consistent, and will need an fsck before that disk state can be remounted), but it will shave off time by not saving memory. If flags includes VIR_DOMAIN_SNAPSHOT_QUIESCE, then this will additionally request that a guest agent quiesce disk state before the live snapshot is taken (increasing the likelihood of a stable disk, rather than a crash-consistent disk; but it requires cooperation from the guest so it is no more reliable than memballoon changes).
As for the XML changes, it makes sense to snapshot just a subset of disks when you only care about crash-consistent state or if you can rely on a guest agent to quiesce the subset of disk(s) you care about, so the existing<domainsnapshot> element needs a new optional subelement to control which disks are snapshotted; additionally, this subelement will be useful for disk image formats that require additional complexity (such as a secondary file name, rather than the inline snapshot feature of qcow2). I'm envisioning something like the following:
<domainsnapshot> <name>whatever</name> <disk name='/path/to/image1' snapshot='no'/> <disk name='/path/to/image2'> <volsnapshot>...</volsnapshot> </disk> </domainsnapshot>
Does it also need to enhance volume's XML so that it can include <volSnapshot>....</volSnapshot>? Since you are already planning to introduce kinds of new storage volume APIs. One might want to manage the snapshots without domain. If so, it might also need modifications on bunch of internal storage volume APIs, such as volDelete, volCreateFromXML, volGetXMLDesc. etc, it might also need to introduce new flags for the according public APIs, the good is that it seems all of these APIs has "flags" argument. Osier
On 07/05/2011 04:26 AM, Osier Yang wrote:
Does it also need to enhance volume's XML so that it can include <volSnapshot>....</volSnapshot>?
No. We _don't_ want to modify virStorageVolGetXMLDesc to output snapshot information from a virStorageVolPtr, rather you would use the new APIs to query whether a virStorageVol has associated snapshot information, and if so, grab a new virStorageVolSnapshotPtr object that wraps the information, and use virStorageVolSnapshotGetXMLDesc on that new object to get the <volSnapshot> XML. That is, this is parallel to the existing virDomainPtr and virDomainSnapshotPtr - the XML for a given domain does not give you any details about snapshots (or checkpoints) tied to the domain; rather, you use the virDomainSnapshot API to get the corresponding snapshot object tied to a domain, then do all your queries on the snapshot object.
Since you are already planning to introduce kinds of new storage volume APIs. One might want to manage the snapshots without domain.
That's the whole point of my RFC. The _existing_ virDomainSnapshot APIs allow management of checkpoints (disk + memory), and can easily be extended to also manage snapshots (disk only) associated with a domain, whereas my proposed new virStorageVolSnapshot API is for management of snapshots of a given disk image without an associated domain.
If so, it might also need modifications on bunch of internal storage volume APIs, such as volDelete, volCreateFromXML, volGetXMLDesc. etc, it might also need to introduce new flags for the according public APIs, the good is that it seems all of these APIs has "flags" argument.
I'm not sure I follow what additional modifications you are envisioning, although I'll give it a shot: Right now, virStorageVolDelete has a flags argument, but does not have any defined flags. What should happen if the virStorageVolPtr has associated snapshots? I can see the use of adding new flags: virStorageVolDelete(,0) - fail if there are associated snapshots virStorageVolDelete(,VIR_STORAGE_VOL_DELETE_SNAPSHOTS) - delete both the image and all snapshots associated with the image I don't see any changes to virStorageVolGetXMLDesc. Right now, virStorageVolCreateFromXML creates a clone storage volume, and takes a flags argument but no defined flags. What happens if the input virStorageVolPtr clonevol passed to that API has associated snapshots? Will the new volume also have the same snapshots as the old volume? Will the cloning also duplicate the snapshots? Is the new volume a flattened version of the input (no backing files in the new image)? There are probably several useful behaviors, and I'm not sure whether they should be controlled by new flags, by new elements in the xmldesc argument, or some combination of those. But you certainly have a point that there is more to think about when cloning an existing volume if that existing volume has associated snapshots. -- Eric Blake eblake@redhat.com +1-801-349-2682 Libvirt virtualization library http://libvirt.org
On Thu, Jun 16, 2011 at 6:41 AM, Eric Blake <eblake@redhat.com> wrote:
Right now, libvirt has a snapshot API via virDomainSnapshotCreateXML, but for qemu domains, it only works if all the guest disk images are qcow2, and qemu rather than libvirt does all the work. However, it has a couple of drawbacks: it is inherently tied to domains (there is no way to manage snapshots of storage volumes not tied to a domain, even though libvirt does that for qcow2 images associated with offline qemu domains by using the qemu-img application). And it necessarily operates on all of the images associated with a domain in parallel - if any disk image is not qcow2, the snapshot fails, and there is no way to select a subset of disks to save. However, it works on both active (disk and memory state) and inactive domains (just disk state).
Hi Eric, Any updates on your proposed snapshot API enhancements? The use case I am particularly interested in is a backup solution using libvirt snapshot APIs to take consistent backups of guests. The two workflows are reading out full snapshots of disk images (full backup) and reading only those blocks that have changed since the last backup (incremental backup). Incremental backups can be done just like full backups except with an API call to get a dirty blocks list. The client only reads out those dirty blocks from the snapshot. Stefan
On 07/27/2011 04:04 AM, Stefan Hajnoczi wrote:
On Thu, Jun 16, 2011 at 6:41 AM, Eric Blake<eblake@redhat.com> wrote:
Right now, libvirt has a snapshot API via virDomainSnapshotCreateXML, but for qemu domains, it only works if all the guest disk images are qcow2, and qemu rather than libvirt does all the work. However, it has a couple of drawbacks: it is inherently tied to domains (there is no way to manage snapshots of storage volumes not tied to a domain, even though libvirt does that for qcow2 images associated with offline qemu domains by using the qemu-img application). And it necessarily operates on all of the images associated with a domain in parallel - if any disk image is not qcow2, the snapshot fails, and there is no way to select a subset of disks to save. However, it works on both active (disk and memory state) and inactive domains (just disk state).
Hi Eric, Any updates on your proposed snapshot API enhancements?
I still need to post a v2 RFC that gives the XML changes needed to support disk snapshots on top of virDomainSnapshotCreateXML. Meanwhile, I still want to add additional API to make it easier to manage offline storage volume snapshots (storage volumes that are not in use by a defined or running domain), although it obviously won't happen by 0.9.4.
The use case I am particularly interested in is a backup solution using libvirt snapshot APIs to take consistent backups of guests. The two workflows are reading out full snapshots of disk images (full backup) and reading only those blocks that have changed since the last backup (incremental backup).
Full backup should be supported via virDomainSnapshotCreateXML, once I have the new XML in place, by using the new qemu snapshot_blkdev monitor commands.
Incremental backups can be done just like full backups except with an API call to get a dirty blocks list. The client only reads out those dirty blocks from the snapshot.
Incremental backups will need more work - qemu does not yet have the monitor commands for exposing which blocks are dirty. Of course, as code is written for qemu, it would be nice to also be thinking about how to support that in libvirt, so once I do propose my XML changes for full snapshots, it would be nice to remember in your review to think about whether it remains extensible to the incremental case. -- Eric Blake eblake@redhat.com +1-801-349-2682 Libvirt virtualization library http://libvirt.org
On Wed, Jul 27, 2011 at 3:17 PM, Eric Blake <eblake@redhat.com> wrote:
On 07/27/2011 04:04 AM, Stefan Hajnoczi wrote:
On Thu, Jun 16, 2011 at 6:41 AM, Eric Blake<eblake@redhat.com> wrote:
Right now, libvirt has a snapshot API via virDomainSnapshotCreateXML, but for qemu domains, it only works if all the guest disk images are qcow2, and qemu rather than libvirt does all the work. However, it has a couple of drawbacks: it is inherently tied to domains (there is no way to manage snapshots of storage volumes not tied to a domain, even though libvirt does that for qcow2 images associated with offline qemu domains by using the qemu-img application). And it necessarily operates on all of the images associated with a domain in parallel - if any disk image is not qcow2, the snapshot fails, and there is no way to select a subset of disks to save. However, it works on both active (disk and memory state) and inactive domains (just disk state).
Hi Eric, Any updates on your proposed snapshot API enhancements?
I still need to post a v2 RFC that gives the XML changes needed to support disk snapshots on top of virDomainSnapshotCreateXML.
Meanwhile, I still want to add additional API to make it easier to manage offline storage volume snapshots (storage volumes that are not in use by a defined or running domain), although it obviously won't happen by 0.9.4.
Previous discussion has focussed on image files. Is part of your plan to extend virStorageBackend and let the LVM backend support the new APIs?
The use case I am particularly interested in is a backup solution using libvirt snapshot APIs to take consistent backups of guests. The two workflows are reading out full snapshots of disk images (full backup) and reading only those blocks that have changed since the last backup (incremental backup).
Full backup should be supported via virDomainSnapshotCreateXML, once I have the new XML in place, by using the new qemu snapshot_blkdev monitor commands.
Excellent :)
Incremental backups can be done just like full backups except with an API call to get a dirty blocks list. The client only reads out those dirty blocks from the snapshot.
Incremental backups will need more work - qemu does not yet have the monitor commands for exposing which blocks are dirty. Of course, as code is written for qemu, it would be nice to also be thinking about how to support that in libvirt, so once I do propose my XML changes for full snapshots, it would be nice to remember in your review to think about whether it remains extensible to the incremental case.
Robert Wang and I would like to help make the incremental use case possible. Like you say, QEMU does not have this feature yet but it makes sense to plan together with libvirt. Will keep you CCed on QEMU discussions about adding this feature. Stefan
participants (9)
-
Daniel P. Berrange -
Daniel Veillard -
Eric Blake -
Jagane Sundar -
Jes Sorensen -
Motonobu Ichimura -
Nicolas Sebrecht -
Osier Yang -
Stefan Hajnoczi