Back to bug 1317450

Who When What Removed Added
Yaniv Lavi 2016-03-14 10:18:03 UTC Assignee sherold amureini
QA Contact gklein ebenahar
Allon Mureinik 2016-03-14 12:27:00 UTC Assignee amureini ishaby
Allon Mureinik 2016-03-16 09:41:44 UTC Depends On 1142762
Summary [RFE] Ability to turn off autoresume of VMs paused due to IO errors [BLOCKED][RFE] Ability to turn off autoresume of VMs paused due to IO errors
Yaniv Kaul 2016-03-27 04:55:41 UTC Summary [BLOCKED][RFE] Ability to turn off autoresume of VMs paused due to IO errors [RFE] Ability to turn off autoresume of VMs paused due to IO errors [blocked on oVirt RFE 1142762 ]
Allon Mureinik 2016-04-03 13:51:46 UTC Target Milestone ovirt-4.0.0 ovirt-4.1.0
Flags ovirt-4.0.0? ovirt-4.1.0?
John Skeoch 2016-04-18 07:34:49 UTC CC yeylon srevivo
Yaniv Kaul 2016-04-28 07:20:53 UTC Summary [RFE] Ability to turn off autoresume of VMs paused due to IO errors [blocked on oVirt RFE 1142762 ] [RFE] Ability to turn off autoresume of VMs paused due to IO errors [blocked on oVirt RFE: bug 1142762 ]
Yaniv Lavi 2016-05-02 10:57:40 UTC Target Milestone ovirt-4.1.0 ovirt-4.0.0-beta
Flags ovirt-4.1.0? ovirt-4.0.0?
Yaniv Kaul 2016-05-05 08:33:40 UTC Alias oVirt_turn_off_autoresume_of_paused_VMs
Yaniv Lavi 2016-05-18 08:55:59 UTC Target Milestone ovirt-4.0.0-beta ovirt-4.0.0-rc
Yaniv Lavi 2016-05-22 06:35:09 UTC Summary [RFE] Ability to turn off autoresume of VMs paused due to IO errors [blocked on oVirt RFE: bug 1142762 ] [RFE] Ability to turn off autoresume of VMs paused due to IO errors
Allon Mureinik 2016-06-07 08:05:17 UTC Assignee ishaby tnisan
Yaniv Kaul 2016-06-15 07:28:40 UTC CC tnisan
Flags needinfo?(tnisan)
Tal Nisan 2016-06-15 07:49:12 UTC Flags needinfo?(tnisan)
Yaniv Kaul 2016-06-15 08:14:36 UTC Target Milestone ovirt-4.0.0-rc ovirt-4.0.2
Flags ovirt-4.0.0? ovirt-4.0.z?
Aharon Canan 2016-06-21 15:23:59 UTC CC acanan
Flags testing_plan_complete-
Yaniv Kaul 2016-07-14 12:14:15 UTC Flags needinfo?(tnisan)
Allon Mureinik 2016-07-17 15:15:37 UTC Target Milestone ovirt-4.0.2 ovirt-4.1.0-alpha
Flags ovirt-4.0.z? ovirt-4.1.0?
Yaniv Lavi 2016-11-03 13:16:43 UTC Target Milestone ovirt-4.1.0-alpha ovirt-4.2.0
Yaniv Lavi 2016-11-03 13:16:52 UTC Flags ovirt-4.1.0? ovirt-4.2.0?
Yaniv Kaul 2017-06-06 19:02:48 UTC Flags needinfo?(amureini)
Allon Mureinik 2017-06-06 21:24:53 UTC Flags needinfo?(tnisan) needinfo?(amureini)
Yaniv Kaul 2017-06-06 21:26:42 UTC Flags needinfo?(amureini)
Allon Mureinik 2017-06-06 21:38:28 UTC Flags needinfo?(amureini)
Yaniv Kaul 2017-06-06 21:50:04 UTC Flags needinfo?(amureini)
Yaniv Lavi 2017-06-11 12:52:08 UTC Summary [RFE] Ability to turn off autoresume of VMs paused due to IO errors [RFE] Have a policy for autoresume of VMs paused due to IO errors (stay paused, turn off, restart with defined time out time)
Allon Mureinik 2017-06-20 13:25:52 UTC Flags needinfo?(amureini)
Allon Mureinik 2017-06-20 13:26:50 UTC CC nsoffer
Flags needinfo?(nsoffer)
Nir Soffer 2017-06-20 17:18:40 UTC CC kwolf
Flags needinfo?(nsoffer) needinfo?(kwolf)
Kevin Wolf 2017-06-21 08:52:19 UTC Flags needinfo?(kwolf)
Yaniv Lavi 2017-07-31 08:36:54 UTC Priority medium high
Yaniv Lavi 2017-09-11 08:28:27 UTC Priority high urgent
Michal Skrivanek 2017-09-19 13:50:59 UTC CC mtessun
Assignee tnisan michal.skrivanek
oVirt Team Storage Virt
Flags needinfo?(mtessun) needinfo?(ylavi)
Martin Tessun 2017-09-22 14:40:57 UTC Flags needinfo?(mtessun)
Michal Skrivanek 2017-09-25 13:43:01 UTC Assignee michal.skrivanek mzamazal
Tomas Jelinek 2017-10-02 06:18:06 UTC Status NEW POST
Link ID oVirt gerrit 82394
Blocks 1417161
CC jcoscia
Link ID Red Hat Knowledge Base (Solution) 3105171
Andrea Perotti 2017-10-02 09:44:17 UTC CC aperotti
Raz Tamir 2017-10-22 08:24:46 UTC Link ID oVirt gerrit 82532
Link ID oVirt gerrit 82533
Link ID oVirt gerrit 82534
Link ID oVirt gerrit 82570
Link ID oVirt gerrit 82571
Link ID oVirt gerrit 82977
Link ID oVirt gerrit 82982
Link ID oVirt gerrit 83029
Link ID oVirt gerrit 83030
Link ID oVirt gerrit 83031
Link ID oVirt gerrit 83032
Link ID oVirt gerrit 83033
CC ratamir
QA Contact ebenahar mavital
meital avital 2017-10-26 10:27:45 UTC Link ID oVirt gerrit 83082
QA Contact mavital pagranat
Flags testing_ack? testing_ack+
Red Hat Bugzilla Rules Engine 2017-10-26 10:27:55 UTC Flags ovirt-4.2? ovirt-4.2+
Raz Tamir 2017-10-28 17:30:55 UTC Link ID oVirt gerrit 83319
CC ratamir
Yaniv Lavi 2017-10-29 17:01:34 UTC Flags needinfo?(ylavi)
Yaniv Lavi 2017-11-02 14:39:10 UTC Flags needinfo?(michal.skrivanek)
Michal Skrivanek 2017-11-02 16:16:37 UTC Flags needinfo?(michal.skrivanek)
Tomas Jelinek 2017-11-08 08:46:43 UTC Status POST MODIFIED
CC tjelinek
Tomas Jelinek 2017-11-08 10:11:53 UTC Doc Text Feature:
Previously, if the VM has been stopped due to IO Error, there was no way how to configure what should happen after the storage gets fixed. The only option was "auto resume", which resumed the VM. This feature adds two more options configurable per VM: "Kill" and "Leave Paused".

Reason:
There are ways how the "auto resume" together with HA VM using VM lease could lead to split brain. Other reason is that it can interfere with custom HA solutions.

Result:
Now the user can configure 3 resume policies for VMs:
- auto resume (the one which used to be the only one)
- leave paused
- kill
Tomas Jelinek 2017-11-08 10:13:15 UTC Doc Text Feature:
Previously, if the VM has been stopped due to IO Error, there was no way how to configure what should happen after the storage gets fixed. The only option was "auto resume", which resumed the VM. This feature adds two more options configurable per VM: "Kill" and "Leave Paused".

Reason:
There are ways how the "auto resume" together with HA VM using VM lease could lead to split brain. Other reason is that it can interfere with custom HA solutions.

Result:
Now the user can configure 3 resume policies for VMs:
- auto resume (the one which used to be the only one)
- leave paused
- kill
Feature:
Previously, if the VM has been paused due to IO Error, there was no way how to configure what should happen after the storage gets fixed. The only option was "auto resume", which resumed the VM. This feature adds two more options configurable per VM: "Kill" and "Leave Paused".

Reason:
There are ways how the "auto resume" together with HA VM using VM lease could lead to split brain. Other reason is that it can interfere with custom HA solutions.

Result:
Now the user can configure 3 resume policies for VMs:
- auto resume (the one which used to be the only one)
- leave paused
- kill
Gil Klein 2017-11-14 07:12:07 UTC Status MODIFIED ON_QA
Sandro Bonazzola 2017-12-06 07:56:53 UTC Target Release --- 4.2.0
Francisco Garcia 2018-01-19 07:51:13 UTC CC fgarciad
Marina Kalinin 2018-02-01 19:29:10 UTC CC mkalinin
Marina Kalinin 2018-02-01 23:18:44 UTC See Also https://bugzilla.redhat.com/show_bug.cgi?id=1540548
Polina 2018-04-30 05:59:19 UTC Flags testing_plan_complete- testing_plan_complete+
Avital Pinnick 2018-05-01 11:21:39 UTC CC apinnick
Doc Text Feature:
Previously, if the VM has been paused due to IO Error, there was no way how to configure what should happen after the storage gets fixed. The only option was "auto resume", which resumed the VM. This feature adds two more options configurable per VM: "Kill" and "Leave Paused".

Reason:
There are ways how the "auto resume" together with HA VM using VM lease could lead to split brain. Other reason is that it can interfere with custom HA solutions.

Result:
Now the user can configure 3 resume policies for VMs:
- auto resume (the one which used to be the only one)
- leave paused
- kill
Previously, when a virtual machine was paused due to IO error, the only resume policy was "Auto Resume", which resumed the virtual machine. "Auto Resume" was problematic because it could interfere with custom HA solutions. In the current release, the "Kill" and "Leave Paused" resume policies have been added. "Leave Paused" was introduced for users who prefer this option because they have their own HA implementation. "Kill" allows virtual machines with a lease to automatically restart on another host in the event of irrecoverable outage.
Polina 2018-05-02 13:24:17 UTC Status ON_QA VERIFIED
Michal Skrivanek 2018-05-03 09:57:43 UTC Doc Text Previously, when a virtual machine was paused due to IO error, the only resume policy was "Auto Resume", which resumed the virtual machine. "Auto Resume" was problematic because it could interfere with custom HA solutions. In the current release, the "Kill" and "Leave Paused" resume policies have been added. "Leave Paused" was introduced for users who prefer this option because they have their own HA implementation. "Kill" allows virtual machines with a lease to automatically restart on another host in the event of irrecoverable outage. Previously, when a virtual machine was paused due to IO error, the only resume policy was "Auto Resume", which resumed the virtual machine. "Auto Resume" was problematic because it could interfere with custom HA solutions. In the current release, the "Kill" and "Leave Paused" resume policies have been added. "Leave Paused" was introduced for users who prefer this option because they have their own HA implementation. "Kill" allows virtual machines with a lease to automatically restart on another host in the event of irrecoverable outage.

Note that the IO Error reporting depends on the underlying storage protocol. In general on FC storage it is fast to detect, but typical default settings for NFS mounts will not cause an IO Error for minutes
Avital Pinnick 2018-05-03 10:14:57 UTC Doc Text Previously, when a virtual machine was paused due to IO error, the only resume policy was "Auto Resume", which resumed the virtual machine. "Auto Resume" was problematic because it could interfere with custom HA solutions. In the current release, the "Kill" and "Leave Paused" resume policies have been added. "Leave Paused" was introduced for users who prefer this option because they have their own HA implementation. "Kill" allows virtual machines with a lease to automatically restart on another host in the event of irrecoverable outage.

Note that the IO Error reporting depends on the underlying storage protocol. In general on FC storage it is fast to detect, but typical default settings for NFS mounts will not cause an IO Error for minutes
Previously, when a virtual machine was paused due to IO error, the only resume policy was "Auto Resume", which resumed the virtual machine. "Auto Resume" was problematic because it could interfere with custom HA solutions. In the current release, the "Kill" and "Leave Paused" resume policies have been added. "Leave Paused" was introduced for users who prefer this option because they have their own HA implementation. "Kill" allows virtual machines with a lease to automatically restart on another host in the event of irrecoverable outage.

The speed of IO Error reporting depends on the underlying storage protocol. On FC storage, IO Errors are generally detected quickly, while on NFS mounts with typical default settings, they may not be detected for several minutes.
Sandro Bonazzola 2018-05-04 10:45:57 UTC Status VERIFIED CLOSED
Resolution --- CURRENTRELEASE
Last Closed 2018-05-04 06:45:57 UTC
Shivraj 2018-11-21 09:36:27 UTC CC shipatil
Yaniv Kaul 2018-11-21 09:55:35 UTC Flags needinfo?(shipatil)
Shivraj 2018-11-21 17:25:21 UTC Flags needinfo?(shipatil)
Yaniv Kaul 2018-11-21 17:30:33 UTC CC ykaul
Gil Klein 2019-04-28 10:42:55 UTC CC gklein

Back to bug 1317450