On 18/01, garcetto wrote:
Good evening, I see two services and both up, but only one seems to do all backup/restore jobs. Any clue? Thx
Hi, I assume config option "backup_use_same_host" is not set in your cinder backup service configuration or if it's set is set to "False". If it's set to "True", then that's the cause. If it's an incremental backup, then it is being handled by the same service because of the bug fix for bug#1952805 [1]. I don't agree with the bug fix, but it is what we currently have. If none of the 2 above is the issue, then then it may just be bad luck, because backup creation requests are randomized to one of the available services [2]. Cheers, Gorka. [1]: https://bugs.launchpad.net/cinder/+bug/1952805 [2]: https://github.com/openstack/cinder/blob/b8cd101fe41066431b3ccf715780e9cc8b3...
Il Gio 18 Gen 2024, 17:39 Alan Bishop <abishop@redhat.com> ha scritto:
On Wed, Jan 17, 2024 at 7:33 AM garcetto <garcetto@gmail.com> wrote:
good afternoon, i was trying cinder-backup, deployed on 2 containers. the first one seems to do all work, the second is sitting idle.
What does "openstack volume service list" report?
You should see two cinder-backup services. If you only see one then perhaps they are each reporting their status identically (same host?), which makes them appear as just a single service.
If you do see both services, are they both "up" ?
Alan
I have a couple of questions: 1. can i parallelize backup jobs so are ran on both containers? 2. in case of first container down, the second seems not to be able to do backups/restores, nay clue?
thank you.