NFS stale file handle / alpine linux

[dsteinkopf]

When trying to upgrade from 6.3.4 to 6.7.0 I have serious problems: Some (but not all) of my mounted docker volumes can not be accessed inside the container because of "NFS stale file handle" errors when accessing the volumes.

Background: My docker volumes lie on a directory which is mounted by the docker host via NFS.

I figured out that the switch to alpine linux starts this issue (saying "starts" because I assume that alpine inside the docker container is not the "real reason" for the problem).

My most important results from my tests:

• Using 6.3.4 (not yet alpine) does NOT have the problem.
• Using 6.6.2 (alpine) DOES have the problem.
• 6.7.0 "patched back" to non-alpine by reversing 7e96bab does NOT have the problem (but another one - I'll open another issue for that).

My questions:

• What was the reason for switching to alpine? (just being curious)
• Would it be possible to always provide a non-alpine version of the images in future? (like other docker image "distributions" do)
  • If not: I would have to do this manually every time a new version of the image is released (as I did for 6.7.0 - see above)
  • If yes: How is that done? Can I do it or help?

[dsteinkopf]

problem solved by upgrading confluence to mysql.