- Devel - libvirt List Archives

[libvirt] libvirt and clustered volume group problems
by Rommer 25 Oct '11

25 Oct '11

Hello, Description of problem: a) libvirt should use vgchange -aly/-aln instead of vgchange -ay/-an for clustered volume groups b) libvirt should skip not activated/suspended volumes during pool-create/pool-refresh for logical pools Version-Release number of selected component (if applicable): libvirt-0.9.6 How reproducible: always Steps to Reproduce: a) vgchange -ay/-an 1. Build 2-node cluster 2. Create clustered volume group 3. Create logical storage pool in libvirt 4. Open/mount any volume on the first node 5. Try to destroy storage pool on the second node b) not activated volumes: 1. Build 2-node cluster 2. Create clustered volume group 3. Activate exclusively any volume on the first node 4. Try to create pool on the second node Actual results: a) vgchange -ay/-an # virsh pool-destroy vg error: Failed to destroy pool vg error: internal error '/sbin/vgchange -an vg' exited with non-zero status 5 and signal 0: Error locking on node node1: LV vg/lv1 in use: not deactivating b) not activated volumes: # virsh pool-create vg.xml error: Failed to create pool from vg.xml error: internal error lvs command failed Expected results: a) vgchange -ay/-an # virsh pool-destroy vg Pool vg destroyed b) not activated volumes: # virsh pool-create vg.xml Pool vg created from vg.xml Additional info: Attached patch fix problem for me. Original bug report: https://bugzilla.redhat.com/show_bug.cgi?id=748437 a) https://bugzilla.redhat.com/show_bug.cgi?id=748248 b) https://bugzilla.redhat.com/show_bug.cgi?id=748282 -- С уважением, Роман Шишнёв, CTO | ActiveCloud | http://www.active.by Т +375 17 2 911 511 доб. 308 | rommer(a)active.by Облачные решения | Серверы и инфраструктура | IaaS | SaaS | Хостинг

2 1

[libvirt] How to use "virsh migrat" with p2p option?
by Guangya Liu 25 Oct '11

25 Oct '11

Hi libvirt support, Can you please give me some example of how to use "virsh migrate --live" with p2p option including both source host and target host? I try to get some info from your website, but no info with migrate. [root@vmoactive02 qemu]# virsh help migrate NAME migrate - migrate domain to another host SYNOPSIS migrate [--live] [--p2p] [--direct] [--tunnelled] [--persistent] [--undefinesource] [--suspend] [--copy-storage-all] [--copy-storage-inc] <domain> <desturi> [<migrateuri>] [<dname>] DESCRIPTION Migrate domain to another host. Add --live for live migration. OPTIONS --live live migration --p2p peer-2-peer migration --direct direct migration --tunnelled tunnelled migration --persistent persist VM on destination --undefinesource undefine VM on source --suspend do not restart the domain on the destination host --copy-storage-all migration with non-shared storage with full disk copy --copy-storage-inc migration with non-shared storage with incremental copy (same base image shared between source and destination) [--domain] <string> domain name, id or uuid [--desturi] <string> connection URI of the destination host [--migrateuri] <string> migration URI, usually can be omitted [--dname] <string> rename to new name during migration (if supported) Thanks, Guangya

2 3

[libvirt] [PATCH] lxc: check for waitpid failure
by Eric Blake 24 Oct '11

24 Oct '11

Detected by Coverity. waitpid() can leak resources if it fails with EINTR, so it should never be used without checking return status. * src/lxc/lxc_container.c (lxcContainerAvailable): Use safer virWaitPid. --- src/lxc/lxc_container.c | 5 ++--- 1 files changed, 2 insertions(+), 3 deletions(-) diff --git a/src/lxc/lxc_container.c b/src/lxc/lxc_container.c index e9891f7..06ccf7e 100644 --- a/src/lxc/lxc_container.c +++ b/src/lxc/lxc_container.c @@ -1229,7 +1229,6 @@ int lxcContainerAvailable(int features) int cpid; char *childStack; char *stack; - int childStatus; if (features & LXC_CONTAINER_FEATURE_USER) flags |= CLONE_NEWUSER; @@ -1251,8 +1250,8 @@ int lxcContainerAvailable(int features) VIR_DEBUG("clone call returned %s, container support is not enabled", virStrerror(errno, ebuf, sizeof ebuf)); return -1; - } else { - waitpid(cpid, &childStatus, 0); + } else if (virPidWait(cpid, NULL) < 0) { + return -1; } VIR_DEBUG("Mounted all filesystems"); -- 1.7.4.4

2 3

[libvirt] [PATCH] qemu: avoid leaking uninit data from hotplug to dumpxml
by Eric Blake 24 Oct '11

24 Oct '11

Detected by Coverity. Both text and JSON monitors set only the bus and unit fields, which means driveAddr.controller spends life as garbage on the stack, and is then memcpy()'d into the in-memory representation which the user can see via dumpxml. * src/qemu/qemu_hotplug.c (qemuDomainAttachSCSIDisk): Only copy defined fields. --- I have to admit that Coverity is good - it took me several minutes to follow the trail down to qemu_monitor_{text,json}.c and prove to myself that driveAddr.controller really is untouched on success. I didn't actually try to exploit this one - it depends on whatever is already on the stack, and your compiler optimization levels, before you would ever see dumpxml giving bogus information in the <address controller='garbage'> field of the hotplugged <disk>. src/qemu/qemu_hotplug.c | 3 ++- 1 files changed, 2 insertions(+), 1 deletions(-) diff --git a/src/qemu/qemu_hotplug.c b/src/qemu/qemu_hotplug.c index bfa524b..037f4aa 100644 --- a/src/qemu/qemu_hotplug.c +++ b/src/qemu/qemu_hotplug.c @@ -505,7 +505,8 @@ int qemuDomainAttachSCSIDisk(struct qemud_driver *driver, /* XXX we should probably validate that the addr matches * our existing defined addr instead of overwriting */ disk->info.type = VIR_DOMAIN_DEVICE_ADDRESS_TYPE_DRIVE; - memcpy(&disk->info.addr.drive, &driveAddr, sizeof(driveAddr)); + disk->info.addr.drive.bus = driveAddr.bus; + disk->info.addr.drive.unit = driveAddr.unit; } } qemuDomainObjExitMonitorWithDriver(driver, vm); -- 1.7.4.4

2 2

[libvirt] [PATCH] Report 'errno' in int1 field of virErrorPtr
by Daniel P. Berrange 24 Oct '11

24 Oct '11

From: "Daniel P. Berrange" <berrange(a)redhat.com> When reporting a system error (VIR_ERR_SYSTEM_ERROR) via virReportSystemError, we should copy the errno value into the 'int1' field of the virErrorPtr struct. This allows callers to detect certain errno conditions & discard the error * src/util/virterror.c: Place errno value in int1 field --- src/util/virterror.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/src/util/virterror.c b/src/util/virterror.c index 5006fa2..a9d4d38 100644 --- a/src/util/virterror.c +++ b/src/util/virterror.c @@ -1327,7 +1327,7 @@ void virReportSystemErrorFull(int domcode, virRaiseErrorFull(filename, funcname, linenr, domcode, VIR_ERR_SYSTEM_ERROR, VIR_ERR_ERROR, - msg, msgDetail, NULL, -1, -1, msg, msgDetail); + msg, msgDetail, NULL, theerrno, -1, msg, msgDetail); errno = save_errno; } -- 1.7.6.4

2 1

[libvirt] [PATCH] Fix off-by-one printing month in logging code
by Daniel P. Berrange 24 Oct '11

24 Oct '11

From: "Daniel P. Berrange" <berrange(a)redhat.com> The field 'mon' in 'struct tm' gives months 0-11, where as humans tend to expect months 1-12. Thus the month number needing adjusting by 1 * src/util/logging.c: Use human friendly month number --- src/util/logging.c | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/src/util/logging.c b/src/util/logging.c index 9df9003..17c7e84 100644 --- a/src/util/logging.c +++ b/src/util/logging.c @@ -628,6 +628,7 @@ virLogFormatTimestamp(void) gettimeofday(&cur_time, NULL); localtime_r(&cur_time.tv_sec, &time_info); time_info.tm_year += 1900; + time_info.tm_mon += 1; if (virAsprintf(&str, "%4d-%02d-%02d %02d:%02d:%02d.%03d", time_info.tm_year, time_info.tm_mon, time_info.tm_mday, -- 1.7.6.4

2 1

[libvirt] [PATCH V2 0/4] Support for multiple IP addresses using lists
by Stefan Berger 24 Oct '11

24 Oct '11

This patch series builds on the previously posted patch series https://www.redhat.com/archives/libvir-list/2011-October/msg00912.html and introduces the capability to assign a list to a variable and have multiple rules instantiated, one for each item in the list. This means, that if for example a variable like IP has been assigned the following values IP = [1.2.3.4, 5.6.7.8, 10.0.0.1] it will generate 3 rules, which then in turn allows us to build filters that can evaluate multiple possible values per field, i.e., allow the filtering for multiple IP addresses (per interface). It would then need David Steven's patch for support of 'RETURN' (and 'CONTINUE') target(s). v2: - reimplementation of iterator - other nits Regards, Stefan

2 6

[libvirt] [PATCH] storage: Fix error message on vol-create-from failure
by Ryota Ozaki 24 Oct '11

24 Oct '11

If vol-create-from is failed due to 'input volume not found', virsh outputs like this: $ sudo virsh vol-create-from testpool test-vol.xml test.img error: failed to get vol 'test.img', specifying --pool might help error: Storage volume not found: no storage vol with matching path However, '--pool' is incorrect because it is already specified as second argument ('testpool' in this case). It should be "--inputpool". The patch fixes this by using pooloptname, which will be "inputpool" in this case and "pool" in other cases, as error message. --- tools/virsh.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index 72344f0..5544a41 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -15216,8 +15216,8 @@ vshCommandOptVolBy(vshControl *ctl, const vshCmd *cmd, if (pool) vshError(ctl, _("failed to get vol '%s'"), n); else - vshError(ctl, _("failed to get vol '%s', specifying --pool " - "might help"), n); + vshError(ctl, _("failed to get vol '%s', specifying --%s " + "might help"), n, pooloptname); } if (pool) -- 1.7.4.1

1 1

[libvirt] [PATCH] qemu: Forbid specify target path for disk vol
by Osier Yang 24 Oct '11

24 Oct '11

target path for disk vol is just meaningless, and it's never used but free()ed silently when creating the volume. --- src/conf/storage_conf.c | 6 ++++++ 1 files changed, 6 insertions(+), 0 deletions(-) diff --git a/src/conf/storage_conf.c b/src/conf/storage_conf.c index e893b2d..5f4f264 100644 --- a/src/conf/storage_conf.c +++ b/src/conf/storage_conf.c @@ -1052,6 +1052,12 @@ virStorageVolDefParseXML(virStoragePoolDefPtr pool, } ret->target.path = virXPathString("string(./target/path)", ctxt); + if (ret->target.path && (pool->type == VIR_STORAGE_POOL_DISK)) { + virStorageReportError(VIR_ERR_XML_ERROR, "%s", + _("target path for disk vol is just meaningless")); + goto cleanup; + } + if (options->formatFromString) { char *format = virXPathString("string(./target/format/@type)", ctxt); if (format == NULL) -- 1.7.6

2 2

[libvirt] Expose qemu monitor command query_block?
by Osier Yang 24 Oct '11

24 Oct '11

Command "query-block" is only used internally for medium's checking before migration currently. Should we expose it as a public API, such as virDomainQueryBlock (or a qemu specific API, virDomainQemuQueryBlock)? It has enough useful infomation which user may care out, such as "io-status", "locked", "removable", "inserted". And perhaps more info in future. Osier

2 2