- Devel - libvirt List Archives

[libvirt] [PATCH] Fix now dead cleanup of VMs on libvirtd restart
by Daniel P. Berrange 15 Jul '11

15 Jul '11

From: "Daniel P. Berrange" <berrange(a)redhat.com> When libvirtd restarts it will attempt to reconnect to existing LXC containers. If it loads a XML state file for the container the container will appear running. If we fail to read the PID file, or fail to connect to the LXC monitor, we should be killing off the guest, but if the VMs cgroup does not exist any more, cleanup will get skipped. Reading the PID file is also pointless since the PID is in the XML statefile In lxcReconnectVM we do not need to read the PID file. If part of the reconnect process fails we need to run the VM terminate code as a safety net. In lxcVMTerminate, if we can't obtain the VM cgroup, we know the process has died, but we must still run lxcVMCleanup to clear out the virDomainObjPtr live state * src/lxc/lxc_driver.c: Fix cleanup of dead VMs on restart --- src/lxc/lxc_driver.c | 60 +++++++++++++++++++++++++------------------------ 1 files changed, 31 insertions(+), 29 deletions(-) diff --git a/src/lxc/lxc_driver.c b/src/lxc/lxc_driver.c index c8d78d5..03360e5 100644 --- a/src/lxc/lxc_driver.c +++ b/src/lxc/lxc_driver.c @@ -1340,22 +1340,26 @@ static int lxcVmTerminate(lxc_driver_t *driver, return -1; } - if (virCgroupForDomain(driver->cgroup, vm->def->name, &group, 0) != 0) - return -1; - - rc = virCgroupKillPainfully(group); - if (rc < 0) { - virReportSystemError(-rc, "%s", - _("Failed to kill container PIDs")); - rc = -1; - goto cleanup; - } - if (rc == 1) { - lxcError(VIR_ERR_INTERNAL_ERROR, "%s", - _("Some container PIDs refused to die")); - rc = -1; - goto cleanup; + if (virCgroupForDomain(driver->cgroup, vm->def->name, &group, 0) == 0) { + rc = virCgroupKillPainfully(group); + if (rc < 0) { + virReportSystemError(-rc, "%s", + _("Failed to kill container PIDs")); + rc = -1; + goto cleanup; + } + if (rc == 1) { + lxcError(VIR_ERR_INTERNAL_ERROR, "%s", + _("Some container PIDs refused to die")); + rc = -1; + goto cleanup; + } + } else { + /* If cgroup doesn't exist, the VM pids must have already + * died and so we're just cleaning up stale state + */ } + lxcVmCleanup(driver, vm, reason); rc = 0; @@ -2172,32 +2176,24 @@ lxcReconnectVM(void *payload, const void *name ATTRIBUTE_UNUSED, void *opaque) lxcDomainObjPrivatePtr priv; virDomainObjLock(vm); + VIR_DEBUG("Reconnect %d %d %d\n", vm->def->id, vm->pid, vm->state.state); priv = vm->privateData; - if ((priv->monitor = lxcMonitorClient(driver, vm)) < 0) { - goto cleanup; - } - - /* Read pid from controller */ - if ((virFileReadPid(lxc_driver->stateDir, vm->def->name, &vm->pid)) != 0) { - VIR_FORCE_CLOSE(priv->monitor); - goto cleanup; - } if (vm->pid != 0) { vm->def->id = vm->pid; virDomainObjSetState(vm, VIR_DOMAIN_RUNNING, VIR_DOMAIN_RUNNING_UNKNOWN); + if ((priv->monitor = lxcMonitorClient(driver, vm)) < 0) + goto error; + if ((priv->monitorWatch = virEventAddHandle( priv->monitor, VIR_EVENT_HANDLE_ERROR | VIR_EVENT_HANDLE_HANGUP, lxcMonitorEvent, - vm, NULL)) < 0) { - lxcVmTerminate(driver, vm, VIR_DOMAIN_SHUTOFF_FAILED); - virDomainAuditStop(vm, "failed"); - goto cleanup; - } + vm, NULL)) < 0) + goto error; } else { vm->def->id = -1; VIR_FORCE_CLOSE(priv->monitor); @@ -2205,6 +2201,12 @@ lxcReconnectVM(void *payload, const void *name ATTRIBUTE_UNUSED, void *opaque) cleanup: virDomainObjUnlock(vm); + return; + +error: + lxcVmTerminate(driver, vm, VIR_DOMAIN_SHUTOFF_FAILED); + virDomainAuditStop(vm, "failed"); + goto cleanup; } -- 1.7.6

2 1

[libvirt] [PATCH] flags: fix domain_conf migration regression
by Eric Blake 15 Jul '11

15 Jul '11

Commit 461e0f1a broke migration, because there was a code path that tried to enable an internal flag while still going through the public function. Split the internal flag into a separate callback, and validate that flags do not overlap. * src/conf/domain_conf.c (virDomainDefFormat): Split... (virDomainDefFormatInternal): ...to separate the flag check. (virDomainObjFormat): Adjust caller. --- Two regression in one day; my track record is not good right now. I guess this goes to show I didn't test a migration :( src/conf/domain_conf.c | 28 +++++++++++++++++++++------- 1 files changed, 21 insertions(+), 7 deletions(-) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 788981f..8c3e44e 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -9615,8 +9615,18 @@ virDomainHostdevDefFormat(virBufferPtr buf, } -char *virDomainDefFormat(virDomainDefPtr def, - unsigned int flags) +#define DUMPXML_FLAGS \ + (VIR_DOMAIN_XML_SECURE | \ + VIR_DOMAIN_XML_INACTIVE | \ + VIR_DOMAIN_XML_UPDATE_CPU) + +verify((VIR_DOMAIN_XML_INTERNAL_STATUS & DUMPXML_FLAGS) == 0); + +/* This internal version can accept VIR_DOMAIN_XML_INTERNAL_STATUS, + * whereas the public version cannot. */ +static char * +virDomainDefFormatInternal(virDomainDefPtr def, + unsigned int flags) { virBuffer buf = VIR_BUFFER_INITIALIZER; unsigned char *uuid; @@ -9624,9 +9634,7 @@ char *virDomainDefFormat(virDomainDefPtr def, const char *type = NULL; int n, allones = 1; - virCheckFlags(VIR_DOMAIN_XML_SECURE | - VIR_DOMAIN_XML_INACTIVE | - VIR_DOMAIN_XML_UPDATE_CPU, NULL); + virCheckFlags(DUMPXML_FLAGS | VIR_DOMAIN_XML_INTERNAL_STATUS, NULL); if (!(type = virDomainVirtTypeToString(def->virtType))) { virDomainReportError(VIR_ERR_INTERNAL_ERROR, @@ -10055,6 +10063,13 @@ char *virDomainDefFormat(virDomainDefPtr def, return NULL; } +char * +virDomainDefFormat(virDomainDefPtr def, unsigned int flags) +{ + virCheckFlags(DUMPXML_FLAGS, NULL); + return virDomainDefFormatInternal(def, flags); +} + static char *virDomainObjFormat(virCapsPtr caps, virDomainObjPtr obj, @@ -10082,8 +10097,7 @@ static char *virDomainObjFormat(virCapsPtr caps, ((caps->privateDataXMLFormat)(&buf, obj->privateData)) < 0) goto error; - if (!(config_xml = virDomainDefFormat(obj->def, - flags))) + if (!(config_xml = virDomainDefFormatInternal(obj->def, flags))) goto error; virBufferAdd(&buf, config_xml, strlen(config_xml)); -- 1.7.4.4

1 1

[libvirt] [PATCH 1/1] Repoint main page links to libvirt driver pages
by Dave Allan 15 Jul '11

15 Jul '11

From: David Allan <dallan(a)redhat.com> The "libvirt supports:" section on the main page of libvirt.org contains a list of hypervisors with links that point to the sites of the underlying virt technologies. The entry for KVM points to http://www.linux-kvm.org/, for example. People coming to libvirt.org for the first time are likely to know about those sites, and they're probably interested in how libvirt manages those technolgies. This patch points those links to the libvirt driver pages instead. It also consolidates KVM and QEMU as there is only one libvirt driver page for them. Finally, it adds a line about networking support. Dave --- docs/index.html.in | 20 ++++++++++---------- 1 files changed, 10 insertions(+), 10 deletions(-) diff --git a/docs/index.html.in b/docs/index.html.in index 64eb84d..536e354 100644 --- a/docs/index.html.in +++ b/docs/index.html.in @@ -35,32 +35,32 @@ <ul> <li> - The <a href="http://www.cl.cam.ac.uk/Research/SRG/netos/xen/index.html">Xen</a> hypervisor - on Linux and Solaris hosts. + The <a href="http://libvirt.org/drvqemu.html">KVM/QEMU</a> Linux hypervisor </li> <li> - The <a href="http://wiki.qemu.org/Index.html">QEMU</a> emulator + The <a href="http://libvirt.org/drvxen.html">Xen</a> hypervisor + on Linux and Solaris hosts. </li> <li> - The <a href="http://www.linux-kvm.org/">KVM</a> Linux hypervisor + The <a href="http://libvirt.org/drvlxc.html">LXC</a> Linux container system </li> <li> - The <a href="http://lxc.sourceforge.net/">LXC</a> Linux container system + The <a href="http://libvirt.org/drvopenvz.html">OpenVZ</a> Linux container system </li> <li> - The <a href="http://openvz.org/">OpenVZ</a> Linux container system + The <a href="http://libvirt.org/drvuml.html">User Mode Linux</a> paravirtualized kernel </li> <li> - The <a href="http://user-mode-linux.sourceforge.net/">User Mode Linux</a> paravirtualized kernel + The <a href="http://libvirt.org/drvvbox.html">VirtualBox</a> hypervisor </li> <li> - The <a href="http://www.virtualbox.org/">VirtualBox</a> hypervisor + The <a href="http://libvirt.org/drvesx.html">VMware ESX and GSX</a> hypervisors </li> <li> - The <a href="http://www.vmware.com/">VMware ESX and GSX</a> hypervisors + The <a href="http://libvirt.org/drvvmware.html">VMware Workstation and Player</a> hypervisors </li> <li> - The <a href="http://www.vmware.com/">VMware Workstation and Player</a> hypervisors + Virtual networks using bridging, NAT, VEPA and VN-LINK. </li> <li> Storage on IDE/SCSI/USB disks, FibreChannel, LVM, iSCSI, NFS and filesystems -- 1.7.4.4

2 3

[libvirt] [PATCH v3 1/2] vshCommandOptString returns -1 if option is empty and not VSH_OFLAG_EMPTY_OK
by Hu Tao 15 Jul '11

15 Jul '11

Pointed out by Eric. Thanks. --- tools/virsh.c | 9 +++++++-- 1 files changed, 7 insertions(+), 2 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index b7cea58..b43af70 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -12375,8 +12375,13 @@ vshCommandOptString(const vshCmd *cmd, const char *name, const char **value) vshError(NULL, _("Missing required option '%s'"), name); ret = -1; } else { - /* Treat "--option ''" as if option had not been specified. */ - ret = 0; + /* --option '' */ + if (arg->def->flag & VSH_OFLAG_EMPTY_OK) { + ret = 0; + } else { + vshError(NULL, _("option '%s' is empty"), name); + ret = -1; + } } } -- 1.7.5.1

2 4

[libvirt] [PATCH 1/1] Fix error code for storage operations
by Dave Allan 15 Jul '11

15 Jul '11

Many volume operations will fail if the volume in question is being allocated. These operations were returning VIR_ERR_INTERNAL_ERROR when they should be returning VIR_ERR_OPERATION_INVALID. --- src/storage/storage_driver.c | 8 ++++---- 1 files changed, 4 insertions(+), 4 deletions(-) diff --git a/src/storage/storage_driver.c b/src/storage/storage_driver.c index 4f35be0..997b876 100644 --- a/src/storage/storage_driver.c +++ b/src/storage/storage_driver.c @@ -1486,7 +1486,7 @@ storageVolumeCreateXMLFrom(virStoragePoolPtr obj, } if (origvol->building) { - virStorageReportError(VIR_ERR_INTERNAL_ERROR, + virStorageReportError(VIR_ERR_OPERATION_INVALID, _("volume '%s' is still being allocated."), origvol->name); goto cleanup; @@ -1667,7 +1667,7 @@ storageVolumeUpload(virStorageVolPtr obj, } if (vol->building) { - virStorageReportError(VIR_ERR_INTERNAL_ERROR, + virStorageReportError(VIR_ERR_OPERATION_INVALID, _("volume '%s' is still being allocated."), vol->name); goto out; @@ -1876,7 +1876,7 @@ storageVolumeWipe(virStorageVolPtr obj, } if (vol->building) { - virStorageReportError(VIR_ERR_INTERNAL_ERROR, + virStorageReportError(VIR_ERR_OPERATION_INVALID, _("volume '%s' is still being allocated."), vol->name); goto out; @@ -1936,7 +1936,7 @@ storageVolumeDelete(virStorageVolPtr obj, } if (vol->building) { - virStorageReportError(VIR_ERR_INTERNAL_ERROR, + virStorageReportError(VIR_ERR_OPERATION_INVALID, _("volume '%s' is still being allocated."), vol->name); goto cleanup; -- 1.7.4.4

3 3

[libvirt] [PATCH 0/3] RFC: setvcpus: support --current option
by Taku Izumi 15 Jul '11

15 Jul '11

Hi all, This patchset adds the --current option to "virsh setvcpus" command. Currently "virsh setvcpus" command supports "--live" and "--config" , but "--current" option. >From view of consistency, it's reasonable to support "--current" option too. *[PATCH 1/3] setvcpus: extend virDomainSetVcpusFlags API to support current flag *[PATCH 2/3] setvcpus: extend qemuDomainSetVcpusFlags() to support current flag *[PATCH 3/3] setvcpus: add "--current" option to "virsh setvcpus" Best regards, Taku Izumi <izumi.taku(a)jp.fujitsu.com>

2 7

[libvirt] [PATCH]: Add another portion virStream bindings (C# bindings, libvirt-csharp.git)
by Anthony Fox 15 Jul '11

15 Jul '11

2 1

[libvirt] [PATCH 0/8] Introduce new API virDomainUndefineFlags
by Osier Yang 15 Jul '11

15 Jul '11

Hi, Per discussion on https://www.redhat.com/archives/libvir-list/2011-July/msg00556.html, This patch series introduce new API virDomainUndefineFlags, which only support flag VIR_DOMAIN_UNDEFINE_MANAGED_STATE now, might introduce more flags in future. If the flag is specified, the domain managed state file will be removed along with domain undefine process once it exists, the entire domain undefine process will fail if it fails on removing the managed state file. [PATCH 1/8] is small fix on rpc generator scripts, not related with the new API. [PATCH 1/8] Small fixes on rpc generator scripts [PATCH 2/8] UndefineFlags: Define the public API [PATCH 3/8] UndefineFlags: Implement the public API [PATCH 4/8] UndefineFlags: Define the internal API [PATCH 5/8] UndefineFlags: Implement qemu driver handling [PATCH 6/8] UndefineFlags: Implement libxl driver handling [PATCH 7/8] UndefineFlags: Implement the remote protocol [PATCH 8/8] UndefineFlags: Extend virsh undefine to support new flag Regards Osier

4 26

[libvirt] [PATCH] Fix reporting of cert validation failures
by Daniel P. Berrange 15 Jul '11

15 Jul '11

If the server succesfully validates the client cert, it will send back a single byte, under TLS. If it fails, it will close the connection. In this case, we were just reporting the standard I/O error. The original RPC code had a special case hack for the GNUTLS_E_UNEXPECTED_PACKET_LENGTH error code to make us report a more useful error message * src/rpc/virnetclient.c: Return ENOMSG if we get GNUTLS_E_UNEXPECTED_PACKET_LENGTH * src/rpc/virnettlscontext.c: Report cert failure if we see ENOMSG --- src/rpc/virnetclient.c | 2 +- src/rpc/virnettlscontext.c | 3 +++ 2 files changed, 4 insertions(+), 1 deletions(-) diff --git a/src/rpc/virnetclient.c b/src/rpc/virnetclient.c index b9f0fc8..d3965c6 100644 --- a/src/rpc/virnetclient.c +++ b/src/rpc/virnetclient.c @@ -348,7 +348,7 @@ int virNetClientSetTLSSession(virNetClientPtr client, ignore_value(pthread_sigmask(SIG_BLOCK, &oldmask, NULL)); len = virNetTLSSessionRead(client->tls, buf, 1); - if (len < 0) { + if (len < 0 && errno != ENOMSG) { virReportSystemError(errno, "%s", _("Unable to read TLS confirmation")); goto error; diff --git a/src/rpc/virnettlscontext.c b/src/rpc/virnettlscontext.c index b778550..4963ab7 100644 --- a/src/rpc/virnettlscontext.c +++ b/src/rpc/virnettlscontext.c @@ -932,6 +932,9 @@ ssize_t virNetTLSSessionWrite(virNetTLSSessionPtr sess, case GNUTLS_E_INTERRUPTED: errno = EINTR; break; + case GNUTLS_E_UNEXPECTED_PACKET_LENGTH: + errno = ENOMSG; + break; default: errno = EIO; break; -- 1.7.4.4

2 1

[libvirt] [PATCH] Unregister event callback if a fatal error occurs during dispatch
by Daniel P. Berrange 15 Jul '11

15 Jul '11

If we get an I/O error in the async event callback for an RPC client, we might not have consumed all pending data off the wire. This could result in the callback being immediately invoked again. At which point the same I/O might occur. And we're invoked again. And again...And again... Unregistering the async event callback if an error occurs is a good safety net. The real error will be seen when the next RPC method is invoked * src/rpc/virnetclient.c: Unregister event callback on error --- src/rpc/virnetclient.c | 6 ++++-- 1 files changed, 4 insertions(+), 2 deletions(-) diff --git a/src/rpc/virnetclient.c b/src/rpc/virnetclient.c index 9fb8fd4..58d4274 100644 --- a/src/rpc/virnetclient.c +++ b/src/rpc/virnetclient.c @@ -1124,8 +1124,10 @@ void virNetClientIncomingEvent(virNetSocketPtr sock, goto done; } - if (virNetClientIOHandleInput(client) < 0) - VIR_DEBUG("Something went wrong during async message processing"); + if (virNetClientIOHandleInput(client) < 0) { + VIR_WARN("Something went wrong during async message processing"); + virNetSocketRemoveIOCallback(sock); + } done: virNetClientUnlock(client); -- 1.7.4.4

2 1