[Gluster-devel] GlusterFS unify mountpoint, if one of the bricks offline

NovA av.nova at gmail.com
Fri Feb 29 13:51:55 UTC 2008


Hi again!

Little more info to the issue.
As it turned out, the situation described in my previous post also
observed when every brick is online. :/

Just in case, here it is more complete description of my config... I
use glusterFS to provide shared storage for compute cluster (22
compute nodes + 1 head). Each cluster node export it's local disk, and
each mounts GlusterFS. But it's little asymmetric: head node mounts
unify-ALU, compute nodes unify-NUFA. Exact specs:
head node (with namespace): http://glusterfs.pastebin.com/m362afd23
compute nodes: http://glusterfs.pastebin.com/m54f3ad75
The specs are generated and copied to nodes by shell script, so there
should not be any misprints.

WBR,
 Andrey

2008/2/29, NovA <av.nova at gmail.com>:
>  The error like "[fuse-bridge.c:436:fuse_entry_cbk] glusterfs-fuse:
>  264935: / => -1 (116)" appears whenever I touch any glusterFS subdirs.
>  For example, each of the following commands
>  cd /home;  cd Shared;  cd tmp
>  results in the error mentioned. Note, that it always refers to the
>  '/'. Despite the error the commands works as expected, but with
>  noticeable lag.





More information about the Gluster-devel mailing list