Bonnie confuses arla

[Tomas Olsson]

Friedrich Delgado Friedrichs <delgado at dfn-cert.de> writes:

> However bonnie leaves no files and the partition is still full, arla
> seems to be ignorant about that.
> 
On my machine, bonnie does leave a file roughly the size of my cache. And
it is purged from the cache afterwards. This appears to hold for the cat
case as well. Hmm, I may have pressed ^C after the error when running
cat. Anyway, in all cases the test "hangs" for a while as the file is saved
(after the error + close), and after that all seems to be well.

Could you do 
  fs arladeb alm ; fs nnpfsdeb alm
before the bonnie gets "No space..." and send me the resulting logs,
including the time from just before the write failed, and until bonnie has
exited?

( fs arladeb -alm ; fs nnpfsdeb -alm afterwards is probably nice)

> Is there some way for the user to mitigate the full partition effect?
>
Not off the top of my head. One could possibly extend the protocol to pass
cache quota information from arlad to nnpfs, but it's probably not trivial
to get that right.

> The user should not be able to screw the client up in such a way that a
> reboot (or restart of a vital daemon) is needed.
>
Very true. I'll try some more to see if I can reproduce it or just get some
ideas.

> Also the oops seems very difficult to reproduce. I'm not able to
> devote much time to this problem, but if I manage to get some more
> information, I will notify this list.
>
Thanks.

> Test strange-other-characters FAILED
>
Accesses stacken.

> Test checkpwd FAILED
> 
Expected.

thanks
        /Tomas