David Caldwell <[email protected]> writes:
> After dealing with a case for a customer about backing up etcd, I have
> a few questions about the documentation.
>
> 1. Why does it say in the documentation to stop etcd services before
> taking a backup?
> (https://docs.openshift.com/container-platform/3.5/admin_guide/backup_restore.html)
>
> After speaking to various OpenShift engineers, it seems that there is
> a unanimous feeling that taking a backup from etcd live or "hot" works
> fine -- or, if you're really worried, stop just one etcd node and take
> a backup on that node. Stopping all etcd services may generate
> unnecessary downtime.
>
> I also asked the same question on a Google etcd list and the answer
> from one of the CoreOS guys was 'no' (at least for etcd version 3).
> (https://groups.google.com/forum/#!searchin/etcd-dev/backup%7Csort:relevance/etcd-dev/JGGCYhy7N2o/F6hmpD4WAgAJ)
>
> 2. In the same documentation mentioned in my Q1, it does not specify
> whether to stop etcd if you only have one node running it. I feel that
> this is confusing.

just a short note what we have experienced so far:

we are currently running oscp 3.3 with etcd2 and doing regular backups
(once a night) via 'etcdctl backup'. so we do _not_ stop a etcd node or
the hole cluster.

we also had to restore one cluster from these backups, the etcd restore
was not a problem.

ymmv
toni

_______________________________________________
dev mailing list
[email protected]
http://lists.openshift.redhat.com/openshiftmm/listinfo/dev

Reply via email to