Frank Filz

OK, of you're getting that with V5.5.1 we need some debug. A refcount on a fsal_fd is going below 0, so there is an extra unref. Please add this to...

Yea, we need the log with FSAL at FULL_DEBUG leading up to the crash.

Try this patch: https://review.gerrithub.io/c/ffilz/nfs-ganesha/+/1168733?usp=search

Could you possibly try recreating on a local filesystem with FSAL_VFS? Our resources for Gluster support are limited and it would be good to understand if this is a generic...

OK, definitely Gluster based. Does anyone have the bandwidth to play with the code and possibly submit a fix? We're trying to evaluate how much support for FSAL_GLUSTER we can...

> > OK, definitely Gluster based. Does anyone have the bandwidth to play with the code and possibly submit a fix? We're trying to evaluate how much support for FSAL_GLUSTER...

Could you submit a patch to gerrithub following src/CONTRIBUTING_HOWTO.txt

I'm not familiar with using openshift in this way. Ganesha needs to have a filesystem ID as part of it's file handle in order to properly support all the iterations...

Interesting. Both xfs and ext4 are easily handled by FSAL_VFS (there is also a FSAL_XFS which is mostly the same as FSAL_VFS but uses an xfs specific ioctl to do...

> Hi, i am using Ganesha Version 5.7 and still experience this issue. What can i do? I already Posted a question [here on Github](https://github.com/gluster/glusterfs/issues/4268) an on [serverfault](https://serverfault.com/questions/1148345/glusterfs-critical-error-negative-fsal-fd-global-counter-freezes-nfs-shares-s?noredirect=1#comment1499188_1148345). If you...