## Description of problem: Recently, customers have started requesting that disks (particularly direct-attach) propagate errors rather than pause the VM. We have provided steps to do this via a script, and there is an RFE to allow it via the GUI [1] Today I have noticed that if you change the disk properties (ie, 'Enable SCSI Pass-Through'), it resets propagate_errors back to it's default value ('Off'). This probably would not be noticed by the customer initially, until the disk next unexpected had an error or go offline... ## Version-Release number of selected component (if applicable): ovirt-engine-4.2.8.2-0.1.el7ev.noarch (and most likely older versions) ## How reproducible: Always ## Steps to Reproduce: 1. Set a disk to propagate errors, ie: > update base_disks set propagate_errors = 'On' where disk_id = '747ce3bb-45d0-40cb-bf93-45c0b869bbcd'; UPDATE 1 > select disk_id, sgio, propagate_errors pe, disk_alias, disk_description from base_disks where disk_alias = 'rhel7-test2_Disk4'; disk_id | sgio | pe | disk_alias | disk_description --------------------------------------+------+----+-------------------+------------------ 747ce3bb-45d0-40cb-bf93-45c0b869bbcd | | On | rhel7-test2_Disk4 | cf5e (1 row) 2. Change the 'Enable SCSI Pass-Through' setting (VM -> disk -> edit, via the GUI) 3. Check database settings again: select disk_id, sgio, propagate_errors pe, disk_alias, disk_description from base_disks where disk_alias = 'rhel7-test2_Disk4'"; disk_id | sgio | pe | disk_alias | disk_description --------------------------------------+------+-----+-------------------+------------------ 747ce3bb-45d0-40cb-bf93-45c0b869bbcd | 1 | Off | rhel7-test2_Disk4 | cf5e (1 row) ## Actual results: propagate_errors changes back to default value ('Off') ## Expected results: propagate_errors should remain to whatever it was set to (in this case, 'On') ## Additional info: Public kbase article on how to change propagate_errors: https://access.redhat.com/solutions/526303 [1] https://bugzilla.redhat.com/show_bug.cgi?id=1314160
Seems like we're explicitly setting propagate errors to off when flushing, Daniel do you recall what's the reason for doing so?
(In reply to Tal Nisan from comment #2) > Seems like we're explicitly setting propagate errors to off when flushing, > Daniel do you recall what's the reason for doing so? I guess it was just for initializing the property, but no longer needed as it's done in BaseDisk ctr. Send a patch to remove it: https://gerrit.ovirt.org/#/c/97402/
WARN: Bug status wasn't changed from MODIFIED to ON_QA due to the following reason: [Found non-acked flags: '{'rhevm-4.3-ga': '?'}', ] For more info please contact: rhv-devops: Bug status wasn't changed from MODIFIED to ON_QA due to the following reason: [Found non-acked flags: '{'rhevm-4.3-ga': '?'}', ] For more info please contact: rhv-devops
1) Added a direct LUN and attached to a VM with SCSI passthrough enabled 2) Updated the disk propagate errors to On: engine=# update base_disks set propagate_errors = 'On' where disk_id = 'ded5395b-41f2-4f92-b2d5-ba20447dbcfc'; UPDATE 1 engine=# select disk_id, sgio, propagate_errors pe, disk_alias, disk_description from base_disks where disk_alias = 'test_Disk1'; disk_id | sgio | pe | disk_alias | disk_description --------------------------------------+------+----+------------+------------------ ded5395b-41f2-4f92-b2d5-ba20447dbcfc | 1 | On | test_Disk1 | 1f6b 3) Updated the disk with SCSI passthrough to disabled: 2019-03-06 12:11:15,680+02 INFO [org.ovirt.engine.core.bll.storage.disk.UpdateVmDiskCommand] (default task-17) [4c116fa6-5ea7-4882-b958-87cfc8d0cc80] Running command: UpdateVmDiskCommand internal: false. Entiti es affected : ID: ded5395b-41f2-4f92-b2d5-ba20447dbcfc Type: DiskAction group EDIT_DISK_PROPERTIES with role type USER, ID: ded5395b-41f2-4f92-b2d5-ba20447dbcfc Type: DiskAction group CONFIGURE_SCSI_GENERIC_IO with role type ADMIN Propagate errors remained On: engine=# select disk_id, sgio, propagate_errors pe, disk_alias, disk_description from base_disks where disk_alias = 'test_Disk1'; disk_id | sgio | pe | disk_alias | disk_description --------------------------------------+------+----+------------+------------------ ded5395b-41f2-4f92-b2d5-ba20447dbcfc | | On | test_Disk1 | 1f6b (1 row) ============== Used: ovirt-engine-4.3.1.1-0.1.el7.noarch vdsm-4.30.9-1.el7ev.x86_64
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://access.redhat.com/errata/RHEA-2019:1085