[Bugs] [Bug 1709959] Gluster causing Kubernetes containers to enter crash loop with 'mkdir ... file exists' error message

[bugzilla at redhat.com]

https://bugzilla.redhat.com/show_bug.cgi?id=1709959



--- Comment #11 from Jeff Bischoff <jeff.bischoff at turbonomic.com> ---
Mohit, thanks for your feedback!

We have been testing solution #2, and so far so good. We haven't rolled that
changed out to all our environments yet, but we haven't had an incident yet on
the environments that do have the health checks disabled.

Regarding solution #1, we appreciate the suggest and will try out multiplexing. 

Just to confirm, what would the expected behavior be if we had multiplexing
turned on and then ran into the same aio_write failure? You say the process
will send a detach even to the specific brick. What happens after that? Will
the brick eventually be automatically reattached? Or will it still require
admin intervention, even with multiplexing?

I'm assuming its not wise to choose both solution #1 and #2? Better to just
pick one?

-- 
You are receiving this mail because:
You are on the CC list for the bug.
You are the assignee for the bug.