Trouble restoring + upgrading to ovirt 4.5 system after host crashed
Good afternoon all, We had a confluence of events hit all at once and need help desperately. Our Ovirt engine system recently crashed and is unrecoverable. Due to a power maintenance event at the data center, 1/3 of our VM's are offline. I have recent backups from the engine created with engine-backup. I installed a clean Centos 9 and followed the directions to install the ovirt-engine . After I restore the backup, the engine-setup fails on the keycloak configuration. *From clean system:* *Install: **(Observe failed scriptlet during install, but rom install still succeeds)* [root@ovirt2 administrator]# dnf install -y ovirt-engine Last metadata expiration check: 2:08:15 ago on Tue 08 Aug 2023 10:11:31 AM CDT. Dependencies resolved. ============================================================================================================================================================= Package Architecture Version Repository Size ============================================================================================================================================================= Installing: ovirt-engine noarch 4.5.4-1.el9 centos-ovirt45 13 M Installing dependencies: SuperLU x86_64 5.3.0-2.el9 epel 182 k (Snip ...) * Running scriptlet: ovirt-vmconsole-1.0.9-1.el9.noarch 60/425Failed to resolve allow statement at /var/lib/selinux/targeted/tmp/modules/400/ovirt_vmconsole/cil:539Failed to resolve AST/usr/sbin/semodule: Failed!* (Snip ...) xmlrpc-common-3.1.3-1.1.el9.noarch xorg-x11-fonts-ISO8859-1-100dpi-7.5-33.el9.noarch zziplib-0.13.71-9.el9.x86_64 Complete! *Engine-restore (no visible issues):* [root@ovirt2 administrator]# engine-backup --mode=restore --log=restore1.log --file=Downloads/engine-2023-08-06.22.00.02.bak --provision-all-databases --restore-permissions Start of engine-backup with mode 'restore' scope: all archive file: Downloads/engine-2023-08-06.22.00.02.bak log file: restore1.log Preparing to restore: - Unpacking file 'Downloads/engine-2023-08-06.22.00.02.bak' Restoring: - Files ------------------------------------------------------------------------------ Please note: Operating system is different from the one used during backup. Current operating system: centos9 Operating system at backup: centos8 Apache httpd configuration will not be restored. You will be asked about it on the next engine-setup run. ------------------------------------------------------------------------------ Provisioning PostgreSQL users/databases: - user 'engine', database 'engine' - user 'ovirt_engine_history', database 'ovirt_engine_history' - user 'ovirt_engine_history_grafana' on database 'ovirt_engine_history' Restoring: - Engine database 'engine' - Cleaning up temporary tables in engine database 'engine' - Updating DbJustRestored VdcOption in engine database - Resetting DwhCurrentlyRunning in dwh_history_timekeeping in engine database - Resetting HA VM status ------------------------------------------------------------------------------ Please note: The engine database was backed up at 2023-08-06 22:00:19.000000000 -0500 . Objects that were added, removed or changed after this date, such as virtual machines, disks, etc., are missing in the engine, and will probably require recovery or recreation. ------------------------------------------------------------------------------ - DWH database 'ovirt_engine_history' - Grafana database '/var/lib/grafana/grafana.db' You should now run engine-setup. Done. [root@ovirt2 administrator]# *Engine-setup :* [root@ovirt2 administrator]# engine-setup [ INFO ] Stage: Initializing [ INFO ] Stage: Environment setup Configuration files: /etc/ovirt-engine-setup.conf.d/10-packaging-jboss.conf, /etc/ovirt-engine-setup.conf.d/10-packaging.conf, /etc/ovirt-engine-setup.conf.d/20-setup-ovirt-post.conf Log file: /var/log/ovirt-engine/setup/ovirt-engine-setup-20230808124501-joveku.log Version: otopi-1.10.3 (otopi-1.10.3-1.el9) [ INFO ] The engine DB has been restored from a backup *[ ERROR ] Failed to execute stage 'Environment setup': Cannot connect to Keycloak database 'ovirt_engine_keycloak' using existing credentials: ovirt_engine_keycloak@localhost:5432*[ INFO ] Stage: Clean up Log file is located at /var/log/ovirt-engine/setup/ovirt-engine-setup-20230808124501-joveku.log [ INFO ] Generating answer file '/var/lib/ovirt-engine/setup/answers/20230808124504-setup.conf' [ INFO ] Stage: Pre-termination [ INFO ] Stage: Termination *[ ERROR ] Execution of setup failed[root@ovirt2 administrator]#* *Engine-cleanup results:* (snip) [ INFO ] Stage: Clean up Log file is located at /var/log/ovirt-engine/setup/ovirt-engine-remove-20230808120445-mj4eef.log [ INFO ] Generating answer file '/var/lib/ovirt-engine/setup/answers/20230808120508-cleanup.conf' [ INFO ] Stage: Pre-termination [ INFO ] Stage: Termination [ INFO ] Execution of cleanup completed successfully [root@cen-90-tmpl administrator]# *Engine backup (restore) results:* [root@ovirt2 administrator]# engine-backup --mode=restore --log=restore1.log --file=Downloads/engine-2023-08-06.22.00.02.bak --provision-all-databases --restore-permissions Start of engine-backup with mode 'restore' scope: all archive file: Downloads/engine-2023-08-06.22.00.02.bak log file: restore1.log Preparing to restore: - Unpacking file 'Downloads/engine-2023-08-06.22.00.02.bak' Restoring: - Files ------------------------------------------------------------------------------ Please note: Operating system is different from the one used during backup. Current operating system: centos9 Operating system at backup: centos8 Apache httpd configuration will not be restored. You will be asked about it on the next engine-setup run. ------------------------------------------------------------------------------ Provisioning PostgreSQL users/databases: - user 'engine', database 'engine' *FATAL: Existing database 'engine' or user 'engine' found and temporary ones created - Please clean up everything and try again* Any advice would be appreciated. *David Johnson*
Hi, On Tue, Aug 8, 2023 at 9:21 PM David Johnson <djohnson@maxistechnology.com> wrote:
Good afternoon all,
We had a confluence of events hit all at once and need help desperately. Our Ovirt engine system recently crashed and is unrecoverable. Due to a power maintenance event at the data center, 1/3 of our VM's are offline.
I have recent backups from the engine created with engine-backup.
How do you run engine-backup for backups? What version? What OS?
I installed a clean Centos 9 and followed the directions to install the ovirt-engine .
After I restore the backup, the engine-setup fails on the keycloak configuration.
*From clean system:*
*Install: **(Observe failed scriptlet during install, but rom install still succeeds)*
[root@ovirt2 administrator]# dnf install -y ovirt-engine Last metadata expiration check: 2:08:15 ago on Tue 08 Aug 2023 10:11:31 AM CDT. Dependencies resolved.
============================================================================================================================================================= Package Architecture Version Repository Size
============================================================================================================================================================= Installing: ovirt-engine noarch 4.5.4-1.el9 centos-ovirt45 13 M Installing dependencies: SuperLU x86_64 5.3.0-2.el9 epel 182 k
(Snip ...)
* Running scriptlet: ovirt-vmconsole-1.0.9-1.el9.noarch
60/425Failed to resolve allow statement at /var/lib/selinux/targeted/tmp/modules/400/ovirt_vmconsole/cil:539Failed to resolve AST/usr/sbin/semodule: Failed!*
This might cause a problem later on, but I do not think it's related to your current issue.
(Snip ...) xmlrpc-common-3.1.3-1.1.el9.noarch xorg-x11-fonts-ISO8859-1-100dpi-7.5-33.el9.noarch zziplib-0.13.71-9.el9.x86_64
Complete!
*Engine-restore (no visible issues):*
[root@ovirt2 administrator]# engine-backup --mode=restore --log=restore1.log --file=Downloads/engine-2023-08-06.22.00.02.bak --provision-all-databases --restore-permissions Start of engine-backup with mode 'restore' scope: all archive file: Downloads/engine-2023-08-06.22.00.02.bak log file: restore1.log Preparing to restore: - Unpacking file 'Downloads/engine-2023-08-06.22.00.02.bak' Restoring: - Files
------------------------------------------------------------------------------ Please note:
Operating system is different from the one used during backup. Current operating system: centos9 Operating system at backup: centos8
I do not think this is the problem, but you might try as well on centos8.
Apache httpd configuration will not be restored. You will be asked about it on the next engine-setup run.
------------------------------------------------------------------------------ Provisioning PostgreSQL users/databases: - user 'engine', database 'engine' - user 'ovirt_engine_history', database 'ovirt_engine_history' - user 'ovirt_engine_history_grafana' on database 'ovirt_engine_history'
Restoring: - Engine database 'engine' - Cleaning up temporary tables in engine database 'engine' - Updating DbJustRestored VdcOption in engine database - Resetting DwhCurrentlyRunning in dwh_history_timekeeping in engine database - Resetting HA VM status
------------------------------------------------------------------------------ Please note:
The engine database was backed up at 2023-08-06 22:00:19.000000000 -0500 .
Objects that were added, removed or changed after this date, such as virtual machines, disks, etc., are missing in the engine, and will probably require recovery or recreation.
------------------------------------------------------------------------------ - DWH database 'ovirt_engine_history' - Grafana database '/var/lib/grafana/grafana.db'
No Keycloak DB restored. I guess it was not backed up, perhaps not even configured.
You should now run engine-setup. Done. [root@ovirt2 administrator]#
*Engine-setup :*
[root@ovirt2 administrator]# engine-setup [ INFO ] Stage: Initializing [ INFO ] Stage: Environment setup Configuration files: /etc/ovirt-engine-setup.conf.d/10-packaging-jboss.conf, /etc/ovirt-engine-setup.conf.d/10-packaging.conf, /etc/ovirt-engine-setup.conf.d/20-setup-ovirt-post.conf Log file: /var/log/ovirt-engine/setup/ovirt-engine-setup-20230808124501-joveku.log Version: otopi-1.10.3 (otopi-1.10.3-1.el9) [ INFO ] The engine DB has been restored from a backup
*[ ERROR ] Failed to execute stage 'Environment setup': Cannot connect to Keycloak database 'ovirt_engine_keycloak' using existing credentials: ovirt_engine_keycloak@localhost:5432*
Despite the above, engine-setup tries to connect to the Keycloak DB. Not sure why. Perhaps check/share also relevant snippets of the setup log. Anyway, I think you should be able to make it not try this, by finding and removing a file under /etc/ovirt-engine/engine.conf.d/ - should be 12-setup-keycloak.conf. I am writing from memory - do not have an engine setup available to check on anymore.
[ INFO ] Stage: Clean up Log file is located at /var/log/ovirt-engine/setup/ovirt-engine-setup-20230808124501-joveku.log [ INFO ] Generating answer file '/var/lib/ovirt-engine/setup/answers/20230808124504-setup.conf' [ INFO ] Stage: Pre-termination [ INFO ] Stage: Termination
*[ ERROR ] Execution of setup failed[root@ovirt2 administrator]#*
*Engine-cleanup results:* (snip)
(snip) is too late here, to see if it cleaned the engine DB (and others).
[ INFO ] Stage: Clean up Log file is located at /var/log/ovirt-engine/setup/ovirt-engine-remove-20230808120445-mj4eef.log [ INFO ] Generating answer file '/var/lib/ovirt-engine/setup/answers/20230808120508-cleanup.conf' [ INFO ] Stage: Pre-termination [ INFO ] Stage: Termination [ INFO ] Execution of cleanup completed successfully [root@cen-90-tmpl administrator]#
*Engine backup (restore) results:*
[root@ovirt2 administrator]# engine-backup --mode=restore --log=restore1.log --file=Downloads/engine-2023-08-06.22.00.02.bak --provision-all-databases --restore-permissions Start of engine-backup with mode 'restore' scope: all archive file: Downloads/engine-2023-08-06.22.00.02.bak log file: restore1.log Preparing to restore: - Unpacking file 'Downloads/engine-2023-08-06.22.00.02.bak' Restoring: - Files
------------------------------------------------------------------------------ Please note:
Operating system is different from the one used during backup. Current operating system: centos9 Operating system at backup: centos8
Apache httpd configuration will not be restored. You will be asked about it on the next engine-setup run.
------------------------------------------------------------------------------ Provisioning PostgreSQL users/databases: - user 'engine', database 'engine'
*FATAL: Existing database 'engine' or user 'engine' found and temporary ones created - Please clean up everything and try again*
Meaning, the engine DB was not cleaned. If it's for production, I recommend reinstalling the OS. If it's just for testing, you can try: 1. Make sure there are no postgresql clients connected (engine, dwh, grafana, etc.) 2. systemctl stop postgresql-server 3. rm -rf /var/lib/pgsql/data Then try again restore and setup. All of this is also from memory, sorry for mistakes if any. Good luck and best regards, -- Didi
Thank you for your response. I powered up the defunct system and confirmed that the keycloak DB was present in the source database, but the engine-backup tool was not backing it up. I was able to recover with the following these steps: 1. start from clean CentOS 9 2. remove a conflicting man page rpm 2. dnf install ovirt-engine, ignore the script error 3. restore the engine database 4. back up the keycloak database with pg_dump, migrate to new engine host, and restore the keycloak database 5. Add entries to the pg_hba for the keycloak database 6. ensure the keycloak user could connect to the keycloak database 7. run engine-setup (no errors) 8. manually add engin.cer to the java keystore We are now running on Version 4.5.4-1.el9 *David Johnson* *Director of Development, Maxis Technology* 844.696.2947 ext 702 (o) | 479.531.3590 (c) <https://www.linkedin.com/in/pojoguy/> <https://maxistechnology.com/wp-content/uploads/vcards/vcard-David_Johnson.vcf> <https://maxistechnology.com/> *Follow us:* <https://www.linkedin.com/company/maxis-tech-inc/>
participantes (2)
-
David Johnson -
Yedidyah Bar David