Bug 1242181

Summary:

An option to do unsafe live migration of VirtualDomain

Product:

Red Hat Enterprise Linux 7

Reporter:

Sergey Urushkin <urusha.v1.0>

Component:

resource-agents

Assignee:

Oyvind Albrigtsen <oalbrigt>

Status:

CLOSED ERRATA

QA Contact:

cluster-qe <cluster-qe>

Severity:

unspecified

Docs Contact:

Priority:

unspecified

Version:

7.1

CC:

agk, cluster-maint, mnovacek, urusha.v1.0

Target Milestone:

Target Release:

---

Hardware:

Unspecified

OS:

Unspecified

Whiteboard:

Fixed In Version:

resource-agents-3.9.5-61.el7

Doc Type:

Bug Fix

Doc Text:

Story Points:

---

Clone Of:

Clones:

1286650 (view as bug list)

Environment:

Last Closed:

2016-11-03 23:57:25 UTC

Type:

Bug

Regression:

---

Mount Type:

---

Documentation:

---

CRM:

Verified Versions:

Category:

---

oVirt Team:

---

RHEL 7.3 requirements from Atomic Host:

Cloudforms Team:

---

Target Upstream Version:

Embargoed:

Bug Depends On:

Bug Blocks:

1286650

Attachments:

Description	Flags
VirtualDomain unsafe live migration patch	none

Description Sergey Urushkin 2015-07-11 22:22:00 UTC

Created attachment 1051014 [details]
VirtualDomain unsafe live migration patch

Description of problem:
There is no option to do unsafe live migration of VirtualDomain resource.
If domain's image is stored on coherent clustered filesystem or DRBD partition,
it's safe to migrate it even if QEMU's cache mode isn't set to "none" (e.g.
writethrough). So, there should be an option for that in resource agent.

Version-Release number of selected component (if applicable):
resource-agents-3.9.5-40.el7_1.4.x86_64

How reproducible:
Always.

Steps to Reproduce:
1. Create a virtual domain with image on GFS2 filesystem or DRBD (protocol C) partition with cache=writethrough.
2. Create a pacemaker resource for it with allow_migrate=true.
3. Try to migrate it.

Actual results:
Live migration fails with: Unsafe migration: Migration may lead to data corruption if disks use cache != none

Expected results:
Live migration succeeds bypassing libvirt restriction (using virsh "--unsafe" option).

Additional info:
Linux-HA bugreport: https://developerbugs.linuxfoundation.org/show_bug.cgi?id=2674
A patch attached.

Comment 3 Oyvind Albrigtsen 2015-10-07 13:04:48 UTC

https://github.com/ClusterLabs/resource-agents/commit/b22411f2a29684b87c612d3c1bf4992889b165f2

Comment 4 Sergey Urushkin 2015-10-09 06:30:06 UTC

Thanks. That patch works for me too and it's much more flexible.

Comment 8 Oyvind Albrigtsen 2016-03-01 13:16:59 UTC

# pcs resource create VM VirtualDomain config=/etc/libvirt/qemu/vm.xml migration_transport=ssh meta allow-migrate=true


Before:
# rpm -q resource-agents
resource-agents-3.9.5-54.el7_2.6.x86_64

# pcs resource move VM host2
# tail -f /var/log/pacemaker.log
Live migration fails with: Unsafe migration: Migration may lead to data corruption if disks use cache != none


After:
# rpm -q resource-agents
resource-agents-3.9.5-61.el7.x86_64

# pcs resource update VM migrate_options="--unsafe"
# pcs resource move VM host2
# tail -f /var/log/pacemaker.log
INFO: vmhost: Starting live migration to host2 (using virsh --connect=qemu:///system --quiet migrate --live --unsafe vmhost qemu+ssh://host2/system ).
INFO: vmhost: live migration to host2 succeeded.

Comment 9 michal novacek 2016-09-09 07:46:35 UTC

I have verified that it is possible to set additional live migration parameters
with resource-agents-3.9.5-81.el7.x86_64

----

Common setup: Setup bridged virtuals that can be live migrated between cluster
nodes and add them as cluster resources. (1) below.

# pcs resource show R-pool-10-34-70-94
 Resource: R-pool-10-34-70-94 (class=ocf provider=heartbeat type=VirtualDomain)
  Attributes: hypervisor=qemu:///system config=/etc/libvirt/qemu/pool-10-34-70-94.xml \
    migration_transport=ssh migration_downtime=1 \
>>  migrate_options=--unsafe
  Meta Attrs: allow-migrate=true priority=100 
  Utilization: cpu=2 hv_memory=1024
  Operations: start interval=0s timeout=120s (R-pool-10-34-70-94-start-interval-0s)
              stop interval=0s timeout=120s (R-pool-10-34-70-94-stop-interval-0s)
              monitor interval=10 timeout=30 (R-pool-10-34-70-94-monitor-interval-10)
              migrate_from interval=0 timeout=120s (R-pool-10-34-70-94-migrate_from-interval-0)
              migrate_to interval=0 timeout=120 (R-pool-10-34-70-94-migrate_to-interval-0)

# pcs resource move  R-pool-10-34-70-94
# grep VirtualDomain /var/log/messages | tail

...
... INFO: pool-10-34-70-94: Starting live migration to kiff-03.cluster-qe.lab.eng.brq.redhat.com (using: virsh --connect=qemu:///system --quiet migrate --live --unsafe pool-10-34-70-94 qemu+ssh://kiff-03.cluster-qe.lab.eng.brq.redhat.com/system ).
... INFO: pool-10-34-70-94: Setting live migration downtime for pool-10-34-70-94 (using: virsh --connect=qemu:///system --quiet migrate-setmaxdowntime pool-10-34-70-94 1).
... INFO: pool-10-34-70-94: live migration to kiff-03.cluster-qe.lab.eng.brq.redhat.com succeeded.
... INFO: Domain pool-10-34-70-94 already stopped


----
>>> (1)
[root@light-02 ~]# pcs config
Cluster Name: STSRHTS19499
Corosync Nodes:
 light-02.cluster-qe.lab.eng.brq.redhat.com kiff-03.cluster-qe.lab.eng.brq.redhat.com
Pacemaker Nodes:
 kiff-03.cluster-qe.lab.eng.brq.redhat.com light-02.cluster-qe.lab.eng.brq.redhat.com

Resources:
 Clone: dlm-clone
  Meta Attrs: interleave=true ordered=true clone-max=2 
  Resource: dlm (class=ocf provider=pacemaker type=controld)
   Operations: start interval=0s timeout=90 (dlm-start-interval-0s)
               stop interval=0s timeout=100 (dlm-stop-interval-0s)
               monitor interval=30s on-fail=fence (dlm-monitor-interval-30s)
 Clone: clvmd-clone
  Meta Attrs: interleave=true ordered=true clone-max=2 
  Resource: clvmd (class=ocf provider=heartbeat type=clvm)
   Attributes: with_cmirrord=1
   Operations: start interval=0s timeout=90 (clvmd-start-interval-0s)
               stop interval=0s timeout=90 (clvmd-stop-interval-0s)
               monitor interval=30s on-fail=fence (clvmd-monitor-interval-30s)
 Clone: shared-group-clone
  Meta Attrs: clone-max=2 interleave=true 
  Group: shared-group
   Resource: shared-vg (class=ocf provider=heartbeat type=LVM)
    Attributes: exclusive=false partial_activation=false volgrpname=shared
    Operations: start interval=0s timeout=30 (shared-vg-start-interval-0s)
                stop interval=0s timeout=30 (shared-vg-stop-interval-0s)
                monitor interval=10 timeout=30 (shared-vg-monitor-interval-10)
   Resource: etc-libvirt (class=ocf provider=heartbeat type=Filesystem)
    Attributes: device=/dev/shared/etc0 directory=/etc/libvirt/qemu fstype=gfs2 options=
    Operations: start interval=0s timeout=60 (etc-libvirt-start-interval-0s)
                stop interval=0s timeout=60 (etc-libvirt-stop-interval-0s)
                monitor interval=30s (etc-libvirt-monitor-interval-30s)
   Resource: images (class=ocf provider=heartbeat type=Filesystem)
    Attributes: device=/dev/shared/images0 directory=/var/lib/libvirt/images fstype=gfs2 options=
    Operations: start interval=0s timeout=60 (images-start-interval-0s)
                stop interval=0s timeout=60 (images-stop-interval-0s)
                monitor interval=30s (images-monitor-interval-30s)
 Resource: R-pool-10-34-70-94 (class=ocf provider=heartbeat type=VirtualDomain)
  Attributes: hypervisor=qemu:///system config=/etc/libvirt/qemu/pool-10-34-70-94.xml migration_transport=ssh migration_downtime=1 migrate_options=--unsafe
  Meta Attrs: allow-migrate=true priority=100 
  Utilization: cpu=2 hv_memory=1024
  Operations: start interval=0s timeout=120s (R-pool-10-34-70-94-start-interval-0s)
              stop interval=0s timeout=120s (R-pool-10-34-70-94-stop-interval-0s)
              monitor interval=10 timeout=30 (R-pool-10-34-70-94-monitor-interval-10)
              migrate_from interval=0 timeout=120s (R-pool-10-34-70-94-migrate_from-interval-0)
              migrate_to interval=0 timeout=120 (R-pool-10-34-70-94-migrate_to-interval-0)

Stonith Devices:
 Resource: fence-light-02 (class=stonith type=fence_ipmilan)
  Attributes: ipaddr=light-02-ilo login=admin passwd=admin pcmk_host_check=static-list pcmk_host_list=light-02.cluster-qe.lab.eng.brq.redhat.com delay=5
  Operations: monitor interval=60s (fence-light-02-monitor-interval-60s)
 Resource: fence-kiff-03 (class=stonith type=fence_ipmilan)
  Attributes: ipaddr=kiff-03-ilo login=admin passwd=admin pcmk_host_check=static-list pcmk_host_list=kiff-03.cluster-qe.lab.eng.brq.redhat.com
  Operations: monitor interval=60s (fence-kiff-03-monitor-interval-60s)
Fencing Levels:

Location Constraints:
  Resource: R-pool-10-34-70-94
    Disabled on: light-02.cluster-qe.lab.eng.brq.redhat.com (score:-INFINITY) (role: Started) (id:cli-ban-R-pool-10-34-70-94-on-light-02.cluster-qe.lab.eng.brq.redhat.com)
  Resource: clvmd-clone
    Disabled on: pool-10-34-70-94 (score:-INFINITY) (id:location-clvmd-clone-pool-10-34-70-94--INFINITY)
    Disabled on: pool-10-34-70-95 (score:-INFINITY) (id:location-clvmd-clone-pool-10-34-70-95--INFINITY)
  Resource: dlm-clone
    Disabled on: pool-10-34-70-94 (score:-INFINITY) (id:location-dlm-clone-pool-10-34-70-94--INFINITY)
    Disabled on: pool-10-34-70-95 (score:-INFINITY) (id:location-dlm-clone-pool-10-34-70-95--INFINITY)
  Resource: shared-group-clone
    Enabled on: light-02.cluster-qe.lab.eng.brq.redhat.com (score:INFINITY) (id:location-shared-group-clone-light-02.cluster-qe.lab.eng.brq.redhat.com-INFINITY)
    Enabled on: kiff-03.cluster-qe.lab.eng.brq.redhat.com (score:INFINITY) (id:location-shared-group-clone-kiff-03.cluster-qe.lab.eng.brq.redhat.com-INFINITY)
    Disabled on: pool-10-34-70-94 (score:-INFINITY) (id:location-shared-group-clone-pool-10-34-70-94--INFINITY)
    Disabled on: pool-10-34-70-95 (score:-INFINITY) (id:location-shared-group-clone-pool-10-34-70-95--INFINITY)
Ordering Constraints:
  start dlm-clone then start clvmd-clone (kind:Mandatory) (id:order-dlm-clone-clvmd-clone-mandatory)
  start clvmd-clone then start shared-group-clone (kind:Mandatory) (id:order-clvmd-clone-shared-group-clone-mandatory)
  start shared-group-clone then start R-pool-10-34-70-94 (kind:Mandatory) (id:order-shared-group-clone-R-pool-10-34-70-94-mandatory)
Colocation Constraints:
  clvmd-clone with dlm-clone (score:INFINITY) (id:colocation-clvmd-clone-dlm-clone-INFINITY)
  shared-group-clone with clvmd-clone (score:INFINITY) (id:colocation-shared-group-clone-clvmd-clone-INFINITY)
Ticket Constraints:

Alerts:
 Alert: forwarder (path=/usr/tests/sts-rhel7.3/pacemaker/alerts/alert_forwarder.py)
  Recipients:
   Recipient: forwarder-recipient (value=http://virt-009.cluster-qe.lab.eng.brq.redhat.com:37676/)

Resources Defaults:
 No defaults set
Operations Defaults:
 No defaults set

Cluster Properties:
 cluster-infrastructure: corosync
 cluster-name: STSRHTS19499
 dc-version: 1.1.15-10.el7-e174ec8
 have-watchdog: false
 last-lrm-refresh: 1473344258
 no-quorum-policy: freeze
 stonith-enabled: true

Quorum:
  Options:

Comment 11 errata-xmlrpc 2016-11-03 23:57:25 UTC

Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://rhn.redhat.com/errata/RHBA-2016-2174.html