At 6pm UTC, we will be briefly resetting the main NFS server for tools and projects to upgrade the tooling around NFS there.  This should cause a brief climb in processes waiting for IO and then that will settle out.

If things don’t operate as expected, rolling back the change could take a few minutes, but that seems unlikely.  Please let us know if there are prolonged issues in the #wikipedia-cloud channel.