7:29 p.m.
Here is V3 of storage fixes, which addresses review comments of v2.
Summary:
Prerna Saxena (2):
Storage: Introduce shadow vol for refresh while the main vol builds.
Storage : Fix cloning of raw, sparse volumes
src/storage/storage_backend.c | 23 ++++++++++-------------
src/storage/storage_driver.c | 19 ++++++++++++-------
2 files changed, 22 insertions(+), 20 deletions(-)
V2 :
http://www.redhat.com/archives/libvir-list/2015-June/msg01052.html
Changelog:
Modified patches 1,2 per review comments.
--
1.8.3.1
--
Prerna Saxena
Linux Technology Centre,
IBM Systems and Technology Lab,
Bangalore, India
7:35 p.m.
New subject: [libvirt] [PATCH v3 1/2] Storage: Introduce shadow vol for refresh while the main vol builds.
Libvirt periodically refreshes all volumes in a storage pool, including
the volumes being cloned.
While cloning a storage volume from parent, we drop pool locks. Subsequent
volume refresh sometimes changes allocation for an ongoing copy, and leads
to corrupt images.
Fix: Introduce a shadow volume that isolates the volume object under refresh
from the base which has a copy ongoing.
Signed-off-by: Prerna Saxena <prerna(a)linux.vnet.ibm.com>
---
src/storage/storage_driver.c | 14 ++++++++++++--
1 file changed, 12 insertions(+), 2 deletions(-)
diff --git a/src/storage/storage_driver.c b/src/storage/storage_driver.c
index 57060ab..41de8df 100644
--- a/src/storage/storage_driver.c
+++ b/src/storage/storage_driver.c
@@ -1898,7 +1898,7 @@ storageVolCreateXMLFrom(virStoragePoolPtr obj,
{
virStoragePoolObjPtr pool, origpool = NULL;
virStorageBackendPtr backend;
- virStorageVolDefPtr origvol = NULL, newvol = NULL;
+ virStorageVolDefPtr origvol = NULL, newvol = NULL, shadowvol = NULL;
virStorageVolPtr ret = NULL, volobj = NULL;
unsigned long long allocation;
int buildret;
@@ -2010,6 +2010,15 @@ storageVolCreateXMLFrom(virStoragePoolPtr obj,
if (backend->createVol(obj->conn, pool, newvol) < 0)
goto cleanup;
+ /* Make a shallow copy of the 'defined' volume definition, since the
+ * original allocation value will change as the user polls 'info',
+ * but we only need the initial requested values
+ */
+ if (VIR_ALLOC(shadowvol) < 0)
+ goto cleanup;
+
+ memcpy(shadowvol, newvol, sizeof(*newvol));
+
pool->volumes.objs[pool->volumes.count++] = newvol;
volobj = virGetStorageVol(obj->conn, pool->def->name, newvol->name,
newvol->key, NULL, NULL);
@@ -2029,7 +2038,7 @@ storageVolCreateXMLFrom(virStoragePoolPtr obj,
virStoragePoolObjUnlock(origpool);
}
- buildret = backend->buildVolFrom(obj->conn, pool, newvol, origvol, flags);
+ buildret = backend->buildVolFrom(obj->conn, pool, shadowvol, origvol, flags);
storageDriverLock();
virStoragePoolObjLock(pool);
@@ -2071,6 +2080,7 @@ storageVolCreateXMLFrom(virStoragePoolPtr obj,
cleanup:
virObjectUnref(volobj);
virStorageVolDefFree(newvol);
+ VIR_FREE(shadowvol);
if (pool)
virStoragePoolObjUnlock(pool);
if (origpool)
--
1.8.3.1
--
Prerna Saxena
Linux Technology Centre,
IBM Systems and Technology Lab,
Bangalore, India
8:50 p.m.
New subject: [libvirt] [PATCH v3 1/2] Storage: Introduce shadow vol for refresh while the main vol builds.
On Fri, Jun 26, 2015 at 05:05:11PM +0530, Prerna Saxena wrote:
Libvirt periodically refreshes all volumes in a storage pool, including
the volumes being cloned.
While cloning a storage volume from parent, we drop pool locks. Subsequent
volume refresh sometimes changes allocation for an ongoing copy, and leads
to corrupt images.
Fix: Introduce a shadow volume that isolates the volume object under refresh
from the base which has a copy ongoing.
Signed-off-by: Prerna Saxena <prerna(a)linux.vnet.ibm.com>
---
src/storage/storage_driver.c | 14 ++++++++++++--
1 file changed, 12 insertions(+), 2 deletions(-)
ACK
diff --git a/src/storage/storage_driver.c
b/src/storage/storage_driver.c
index 57060ab..41de8df 100644
--- a/src/storage/storage_driver.c
+++ b/src/storage/storage_driver.c
@@ -1898,7 +1898,7 @@ storageVolCreateXMLFrom(virStoragePoolPtr obj,
{
virStoragePoolObjPtr pool, origpool = NULL;
virStorageBackendPtr backend;
- virStorageVolDefPtr origvol = NULL, newvol = NULL;
+ virStorageVolDefPtr origvol = NULL, newvol = NULL, shadowvol = NULL;
virStorageVolPtr ret = NULL, volobj = NULL;
unsigned long long allocation;
int buildret;
@@ -2010,6 +2010,15 @@ storageVolCreateXMLFrom(virStoragePoolPtr obj,
if (backend->createVol(obj->conn, pool, newvol) < 0)
goto cleanup;
+ /* Make a shallow copy of the 'defined' volume definition, since the
+ * original allocation value will change as the user polls 'info',
+ * but we only need the initial requested values
+ */
+ if (VIR_ALLOC(shadowvol) < 0)
+ goto cleanup;
+
+ memcpy(shadowvol, newvol, sizeof(*newvol));
+
pool->volumes.objs[pool->volumes.count++] = newvol;
volobj = virGetStorageVol(obj->conn, pool->def->name, newvol->name,
newvol->key, NULL, NULL);
@@ -2029,7 +2038,7 @@ storageVolCreateXMLFrom(virStoragePoolPtr obj,
virStoragePoolObjUnlock(origpool);
}
- buildret = backend->buildVolFrom(obj->conn, pool, newvol, origvol, flags);
+ buildret = backend->buildVolFrom(obj->conn, pool, shadowvol, origvol, flags);
newvol->target.allocation should also use the value from shadowvol.
I have made the adjustment and pushed the patch.
Jan
3:53 p.m.
New subject: [libvirt] [PATCH v3 1/2] Storage: Introduce shadow vol for refresh while the main vol builds.
On Tuesday 30 June 2015 06:20 PM, Ján Tomko wrote:
On Fri, Jun 26, 2015 at 05:05:11PM +0530, Prerna Saxena wrote:
> Libvirt periodically refreshes all volumes in a storage pool, including
> the volumes being cloned.
> While cloning a storage volume from parent, we drop pool locks. Subsequent
> volume refresh sometimes changes allocation for an ongoing copy, and leads
> to corrupt images.
> Fix: Introduce a shadow volume that isolates the volume object under refresh
> from the base which has a copy ongoing.
>
> Signed-off-by: Prerna Saxena <prerna(a)linux.vnet.ibm.com>
> ---
> src/storage/storage_driver.c | 14 ++++++++++++--
> 1 file changed, 12 insertions(+), 2 deletions(-)
>
ACK
> diff --git a/src/storage/storage_driver.c b/src/storage/storage_driver.c
> index 57060ab..41de8df 100644
> --- a/src/storage/storage_driver.c
> +++ b/src/storage/storage_driver.c
> @@ -1898,7 +1898,7 @@ storageVolCreateXMLFrom(virStoragePoolPtr obj,
> {
> virStoragePoolObjPtr pool, origpool = NULL;
> virStorageBackendPtr backend;
> - virStorageVolDefPtr origvol = NULL, newvol = NULL;
> + virStorageVolDefPtr origvol = NULL, newvol = NULL, shadowvol = NULL;
> virStorageVolPtr ret = NULL, volobj = NULL;
> unsigned long long allocation;
> int buildret;
> @@ -2010,6 +2010,15 @@ storageVolCreateXMLFrom(virStoragePoolPtr obj,
> if (backend->createVol(obj->conn, pool, newvol) < 0)
> goto cleanup;
>
> + /* Make a shallow copy of the 'defined' volume definition, since the
> + * original allocation value will change as the user polls 'info',
> + * but we only need the initial requested values
> + */
> + if (VIR_ALLOC(shadowvol) < 0)
> + goto cleanup;
> +
> + memcpy(shadowvol, newvol, sizeof(*newvol));
> +
> pool->volumes.objs[pool->volumes.count++] = newvol;
> volobj = virGetStorageVol(obj->conn, pool->def->name, newvol->name,
> newvol->key, NULL, NULL);
> @@ -2029,7 +2038,7 @@ storageVolCreateXMLFrom(virStoragePoolPtr obj,
> virStoragePoolObjUnlock(origpool);
> }
>
> - buildret = backend->buildVolFrom(obj->conn, pool, newvol, origvol,
flags);
> + buildret = backend->buildVolFrom(obj->conn, pool, shadowvol, origvol,
flags);
>
newvol->target.allocation should also use the value from shadowvol.
I have made the adjustment and pushed the patch.
Jan
Hi Jan,
Thanks.
Could you also take a look at the second patch in this series ?
Regards,
--
Prerna Saxena
Linux Technology Centre,
IBM Systems and Technology Lab,
Bangalore, India
7:43 p.m.
New subject: [libvirt] [PATCH v3 2/2] Storage : Fix cloning of raw, sparse volumes
When virsh vol-clone is attempted on a raw file where capacity > allocation,
the resulting cloned volume has a size that matches the virtual-size of
the parent; in place of matching its actual, disk size.
This patch fixes the cloned disk to have same _allocated_size_ as
the parent file from which it was cloned.
Ref: http://www.redhat.com/archives/libvir-list/2015-May/msg00050.html
Also fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1130739
Signed-off-by: Prerna Saxena <prerna(a)linux.vnet.ibm.com>
---
src/storage/storage_backend.c | 23 ++++++++++-------------
src/storage/storage_driver.c | 5 -----
2 files changed, 10 insertions(+), 18 deletions(-)
diff --git a/src/storage/storage_backend.c b/src/storage/storage_backend.c
index ce59f63..492b344 100644
--- a/src/storage/storage_backend.c
+++ b/src/storage/storage_backend.c
@@ -342,7 +342,7 @@ virStorageBackendCreateBlockFrom(virConnectPtr conn ATTRIBUTE_UNUSED,
goto cleanup;
}
- remain = vol->target.allocation;
+ remain = vol->target.capacity;
if (inputvol) {
int res = virStorageBackendCopyToFD(vol, inputvol,
@@ -397,7 +397,8 @@ createRawFile(int fd, virStorageVolDefPtr vol,
virStorageVolDefPtr inputvol,
bool reflink_copy)
{
- bool need_alloc = true;
+ bool want_sparse = (inputvol &&
+ (inputvol->target.capacity > vol->target.allocation));
int ret = 0;
unsigned long long remain;
@@ -420,10 +421,9 @@ createRawFile(int fd, virStorageVolDefPtr vol,
* to writing zeroes block by block in case fallocate isn't
* available, and since we're going to copy data from another
* file it doesn't make sense to write the file twice. */
- if (vol->target.allocation) {
- if (fallocate(fd, 0, 0, vol->target.allocation) == 0) {
- need_alloc = false;
- } else if (errno != ENOSYS && errno != EOPNOTSUPP) {
+ if (vol->target.allocation && !want_sparse) {
+ if ((fallocate(fd, 0, 0, vol->target.allocation) != 0) &&
+ (errno != ENOSYS && errno != EOPNOTSUPP)) {
ret = -errno;
virReportSystemError(errno,
_("cannot allocate %llu bytes in file
'%s'"),
@@ -433,22 +433,19 @@ createRawFile(int fd, virStorageVolDefPtr vol,
}
#endif
- remain = vol->target.allocation;
+ remain = vol->target.capacity;
if (inputvol) {
/* allow zero blocks to be skipped if we've requested sparse
- * allocation (allocation < capacity) or we have already
- * been able to allocate the required space. */
- bool want_sparse = !need_alloc ||
- (vol->target.allocation < inputvol->target.capacity);
-
+ * allocation (allocation < capacity)
+ */
ret = virStorageBackendCopyToFD(vol, inputvol, fd, &remain,
want_sparse, reflink_copy);
if (ret < 0)
goto cleanup;
}
- if (remain && need_alloc) {
+ if (remain && !want_sparse) {
if (safezero(fd, vol->target.allocation - remain, remain) < 0) {
ret = -errno;
virReportSystemError(errno, _("cannot fill file '%s'"),
diff --git a/src/storage/storage_driver.c b/src/storage/storage_driver.c
index 41de8df..b8ed006 100644
--- a/src/storage/storage_driver.c
+++ b/src/storage/storage_driver.c
@@ -1976,11 +1976,6 @@ storageVolCreateXMLFrom(virStoragePoolPtr obj,
if (newvol->target.capacity < origvol->target.capacity)
newvol->target.capacity = origvol->target.capacity;
- /* Make sure allocation is at least as large as the destination cap,
- * to make absolutely sure we copy all possible contents */
- if (newvol->target.allocation < origvol->target.capacity)
- newvol->target.allocation = origvol->target.capacity;
-
if (!backend->buildVolFrom) {
virReportError(VIR_ERR_NO_SUPPORT,
"%s", _("storage pool does not support"
--
1.8.3.1
--
Prerna Saxena
Linux Technology Centre,
IBM Systems and Technology Lab,
Bangalore, India
9:54 p.m.
New subject: [libvirt] [PATCH v3 2/2] Storage : Fix cloning of raw, sparse volumes
On Fri, Jun 26, 2015 at 05:13:26PM +0530, Prerna Saxena wrote:
When virsh vol-clone is attempted on a raw file where capacity >
allocation,
the resulting cloned volume has a size that matches the virtual-size of
the parent; in place of matching its actual, disk size.
This patch fixes the cloned disk to have same _allocated_size_ as
the parent file from which it was cloned.
Ref: http://www.redhat.com/archives/libvir-list/2015-May/msg00050.html
Also fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1130739
Signed-off-by: Prerna Saxena <prerna(a)linux.vnet.ibm.com>
---
src/storage/storage_backend.c | 23 ++++++++++-------------
src/storage/storage_driver.c | 5 -----
2 files changed, 10 insertions(+), 18 deletions(-)
diff --git a/src/storage/storage_backend.c b/src/storage/storage_backend.c
index ce59f63..492b344 100644
--- a/src/storage/storage_backend.c
+++ b/src/storage/storage_backend.c
@@ -342,7 +342,7 @@ virStorageBackendCreateBlockFrom(virConnectPtr conn
ATTRIBUTE_UNUSED,
goto cleanup;
}
- remain = vol->target.allocation;
+ remain = vol->target.capacity;
if (inputvol) {
int res = virStorageBackendCopyToFD(vol, inputvol,
@@ -397,7 +397,8 @@ createRawFile(int fd, virStorageVolDefPtr vol,
virStorageVolDefPtr inputvol,
bool reflink_copy)
{
- bool need_alloc = true;
+ bool want_sparse = (inputvol &&
+ (inputvol->target.capacity > vol->target.allocation));
int ret = 0;
unsigned long long remain;
@@ -420,10 +421,9 @@ createRawFile(int fd, virStorageVolDefPtr vol,
* to writing zeroes block by block in case fallocate isn't
* available, and since we're going to copy data from another
* file it doesn't make sense to write the file twice. */
- if (vol->target.allocation) {
- if (fallocate(fd, 0, 0, vol->target.allocation) == 0) {
- need_alloc = false;
- } else if (errno != ENOSYS && errno != EOPNOTSUPP) {
+ if (vol->target.allocation && !want_sparse) {
+ if ((fallocate(fd, 0, 0, vol->target.allocation) != 0) &&
If fallocate succeeds here, there's no need to copy zero bytes from the
source.
+ (errno != ENOSYS && errno != EOPNOTSUPP)) {
ret = -errno;
virReportSystemError(errno,
_("cannot allocate %llu bytes in file
'%s'"),
@@ -433,22 +433,19 @@ createRawFile(int fd, virStorageVolDefPtr vol,
}
#endif
- remain = vol->target.allocation;
+ remain = vol->target.capacity;
This breaks sparse file creation on systems without fallocate.
if (inputvol) {
/* allow zero blocks to be skipped if we've requested sparse
- * allocation (allocation < capacity) or we have already
- * been able to allocate the required space. */
- bool want_sparse = !need_alloc ||
- (vol->target.allocation < inputvol->target.capacity);
-
+ * allocation (allocation < capacity)
+ */
ret = virStorageBackendCopyToFD(vol, inputvol, fd, &remain,
want_sparse, reflink_copy);
I'd rahter use !need_alloc instead of want_sparse here, because
sparseness only makes sense for cloning volumes, but the bool affects
creating new volumes as well.
if (ret < 0)
goto cleanup;
}
- if (remain && need_alloc) {
+ if (remain && !want_sparse) {
if (safezero(fd, vol->target.allocation - remain, remain) < 0) {
This calculation assumes that 'remain' contains the bytes remaining to
the target allocation, but after this patch it contains the bytes
remaining to the target capacity.
ret = -errno;
virReportSystemError(errno, _("cannot fill file '%s'"),
I'll send a v4 shortly.
Jan
3354
days inactive
3361
days old
5 comments
2 participants
participants (2)
-
Ján Tomko -
Prerna Saxena