NetBSD-Users archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

GlusterFS questions


I am trying to use NetBSD GlusterFS client against
Debian 10 GlusterFS cluster. Running rsync from local
disk to gluster volume always ends in rsync process
stuck on client. glusterfs 3.12 get volume unmounted and
sometimes drop core, but gluster 8.1 can be used further
after ^C.

Log file have tons of messages like this:
Unable to get config xattr. FOP : 'FXATTROP' failed on gfid b254d694-97e4-4553-98bc-43c50a45d882 [No message available]

And sometimes like this:
Insufficient available children for this request (have 0, need 4). FOP : 'FXATTROP' failed on gfid 3ef06874-6c07-42a2-9cdb-0f27a9114d5e

Googling made me think it is usual for those messages to come

And nothing changes if writing process hangs or when it does not,
because those messages usually written about any file being written
on volume. It seems writing process get overwhelmed with something
on writing lot of small files, but feel ok with large ones.

Isn't it sounds better to use NetBSD on server cluster if
clients are NetBSD mostly?

Is there any way and any reason to disable xattr or it's vital
for GFS functioning? It may be possible to turn it on mount
level, or glusterfs level or even volume level, but I found
not an option to do that.

Sincerely yours,
Dima Veselov
Physics R&D Establishment of Saint-Petersburg University

Home | Main Index | Thread Index | Old Index