Bug 1459678 - Unable to mount after client upgrade from 3.9 to 3.10
Unable to mount after client upgrade from 3.9 to 3.10
Status: NEW
Product: GlusterFS
Classification: Community
Component: fuse (Show other bugs)
3.10
x86_64 Linux
unspecified Severity high
: ---
: ---
Assigned To: bugs@gluster.org
:
Depends On:
Blocks:
  Show dependency treegraph
 
Reported: 2017-06-07 15:05 EDT by David E. Smith
Modified: 2017-06-07 15 EDT (History)
1 user (show)

See Also:
Fixed In Version:
Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of:
Environment:
Last Closed:
Type: Bug
Regression: ---
Mount Type: ---
Documentation: ---
CRM:
Verified Versions:
Category: ---
oVirt Team: ---
RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: ---


Attachments (Terms of Use)
client logs for successfully mounting with 3.9 client (6.78 KB, text/plain)
2017-06-07 15:05 EDT, David E. Smith
no flags Details
client logs for failed mount with 3.10 client (7.32 KB, text/plain)
2017-06-07 15:05 EDT, David E. Smith
no flags Details

  None (edit)
Description David E. Smith 2017-06-07 15:05:19 EDT
Created attachment 1285873 [details]
client logs for successfully mounting with 3.9 client

Description of problem:
I have an existing GlusterFS file system, whose servers have been upgraded to Gluster 3.10.0 (from gluster.org). The clients are currently on 3.9, and all is well. When I try to upgrade the clients to 3.10, though, the mount process terminates prematurely and the file system is not mounted.

The cluster's versions are, I believe, set correctly. From one of the servers:

gluster> volume get all cluster.op-version
Option                                  Value
------                                  -----
cluster.op-version                      30900
gluster> volume get all cluster.max-op-version
Option                                  Value
------                                  -----
cluster.max-op-version                  31000


Version-Release number of selected component (if applicable):
Servers: Gluster 3.10 (3.10.0-1 RPMs from upstream)
Working Clients: Gluster 3.9 (3.9.1-1 RPMs from upstream)
Non-working Clients: Gluster 3.10 (3.10.0-1 RPMs from upstream)


Attaching logs of successful and unsuccessful mounts. This is the part that looks "interesting" to me, less than half a second after the mount command is started:

[2017-06-07 18:09:27.469408] I [fuse-bridge.c:5089:fuse_thread_proc] 0-fuse: unmounting /mnt

[2017-06-07 18:09:27.470229] W [glusterfsd.c:1329:cleanup_and_exit] (-->/lib64/libpthread.so.0(+0x7aa1) [0x7f2f5ff98aa1] -->/usr/sbin/glusterfs(glusterfs_sigwaiter+0xcd) [0x4063fd] -->/usr/sbin/glusterfs(cleanup_and_exit+0x6a) [0x405efa] ) 0-: received signum (15), shutting down
Comment 1 David E. Smith 2017-06-07 15:05 EDT
Created attachment 1285874 [details]
client logs for failed mount with 3.10 client

Note You need to log in before you can comment on or make changes to this bug.