Description of problem: ======================= When set the client-log-level to DEBUG for one volume. Glustershd logs the Debug information for other volumes too in the system Version-Release number of selected component (if applicable): ============================================================= [12/16/11 - 01:19:31 root@hicks ~]# gluster --version glusterfs 3.3.0.4rhs built on Oct 19 2012 09:16:41 (glusterfs-3.3.0.4rhs-34.el6rhs.x86_64) How reproducible: ================= 1/1 Steps to Reproduce: =================== 1. Create two volumes (replicate and distributed-replicate) 2. Set the client-log-level to Debug for replicate volume "gluster volume set rep client-log-level DEBUG" 3. Execute "gluster volume info" to confirm that the client-log-level is only ser for replicate volume(rep) 4. Perform operation like (volume stop/start) on distributed-replicate volume(dist-rep) Actual results: =============== /var/log/glusterfs/glustershd.log shows the Debug info for both the volumes replicate (rep) and distributed-replicate(dist-rep) [2011-12-16 01:16:37.417551] D [client-handshake.c:1407:client_setvolume_cbk] 0-rep-client-1: clnt-lk-version = 1, server-lk-version = 0 [2011-12-16 01:16:37.417597] I [client-handshake.c:1433:client_setvolume_cbk] 0-rep-client-1: Connected to 10.70.34.117:24012, attached to remote volume '/home/test1/brick1'. [2011-12-16 01:16:37.417631] I [client-handshake.c:1445:client_setvolume_cbk] 0-rep-client-1: Server and Client lk-version numbers are not same, reopening the fds [2011-12-16 01:16:37.417680] D [client-handshake.c:1295:client_post_handshake] 0-rep-client-1: No fds to open - notifying all parents child up [2011-12-16 01:16:37.417716] D [client-handshake.c:489:client_set_lk_version] 0-rep-client-1: Sending SET_LK_VERSION [2011-12-16 01:16:37.419504] I [client-handshake.c:453:client_set_lk_version_cbk] 0-rep-client-1: Server lk version = 1 [2011-12-16 01:16:37.420050] D [afr-self-heald.c:969:afr_find_child_position] 0-rep-replicate-0: child rep-client-1 is remote [2011-12-16 01:17:19.423677] D [client-handshake.c:184:client_start_ping] 0-dist-rep-client-0: returning as transport is already disconnected OR there are no frames (0 || 0) [2011-12-16 01:17:19.423794] D [client-handshake.c:184:client_start_ping] 0-dist-rep-client-1: returning as transport is already disconnected OR there are no frames (0 || 0) [2011-12-16 01:17:19.424007] D [client-handshake.c:184:client_start_ping] 0-dist-rep-client-2: returning as transport is already disconnected OR there are no frames (0 || 0) [2011-12-16 01:17:19.424051] D [client-handshake.c:184:client_start_ping] 0-dist-rep-client-3: returning as transport is already disconnected OR there are no frames (0 || 0) [2011-12-16 01:17:19.424092] D [client-handshake.c:184:client_start_ping] 0-rep-client-0: returning as transport is already disconnected OR there are no frames (0 || 0) [2011-12-16 01:17:19.424129] D [client-handshake.c:184:client_start_ping] 0-rep-client-1: returning as transport is already disconnected OR there are no frames (0 || 0) Expected results: ================= It should log the debug information only for the volume(rep) for which the option "client-log-level" is set to debug, not for other volumes (dist-rep) in system. Additional info:
Created attachment 636104 [details] sosreport and glusterfs logs
needs 'log-level' to be volume specific to achieve this. As of now, log-level is process wide, or xlator specific. It would be great if its graph specific too. I propose this as a feature enhancement.
for now, as a fix, not allow log-level option to be written in nfs/self-heald volfile, thus not getting into this issue for now.
The product version of Red Hat Storage on which this issue was reported has reached End Of Life (EOL) [1], hence this bug report is being closed. If the issue is still observed on a current version of Red Hat Storage, please file a new bug report on the current version. [1] https://rhn.redhat.com/errata/RHSA-2014-0821.html