I just reorganized the wikilink NFS service (T301280) and noticed that only a single VM is actually mounting NFS. That kind of defeats the use of a shard file-system... are there future use-cases that require it, or can this work be moved to a cinder volume?
If the latter, please let me know when you've done so so I can tear down the NFS setup :)
Thanks!
| Subject | Repo | Branch | Lines +/- | |
|---|---|---|---|---|
| nfs-mounts: remove wikilink project | operations/puppet | production | +0 -4 |
| Status | Subtype | Assigned | Task | ||
|---|---|---|---|---|---|
| Open | None | T272395 Cloud: reduce NAT exceptions from cloud to production | |||
| Resolved | Andrew | T291405 [NFS] Reduce or eliminate bare-metal NFS servers | |||
| Resolved | Andrew | T301280 Move project-specific NFS mounts onto project-local NFS servers | |||
| Resolved | Andrew | T301646 Does the 'wikilink' project really need NFS? | |||
| Resolved | jsn.sherman | T302401 migrate wikilink backups from nfs to cinder |
To answer the headline question: no
We were previously using nfs to let one system pull in db dumps from another, but we didn't need to do it as often as I thought we would and we don't do it at all now. Really all we're using this for is for backup storage for the single instance, so I think a cinder volume will suffice. If we do need to push data around occasionally, we can just use scp. As requested, we'll let you know when it's done.
@Andrew I've mounted a new cinder volume, copied over the data, and pointed our application at it and verified that it's happy. I also removed the nfs config from hiera for that instance. The NFS share for this project should be ready for teardown. Thanks for bringing this to our attention!
Change 765331 had a related patch set uploaded (by Andrew Bogott; author: Andrew Bogott):
[operations/puppet@production] nfs-mounts: remove wikilink project
Change 765331 merged by Andrew Bogott:
[operations/puppet@production] nfs-mounts: remove wikilink project