Description of problem: On a WA setup, the glusterfsds crash at some random point which might be because of the race. This bug is to just avoid the crash from happening. RCA tracked separately. They crash at: Reading symbols from /usr/sbin/glusterfsd...Reading symbols from /usr/lib/debug/usr/sbin/glusterfsd.debug...done. done. [New LWP 3816] [New LWP 3817] [New LWP 3823] [New LWP 3813] [New LWP 3814] [New LWP 3815] [New LWP 3812] warning: Could not load shared library symbols for /lib64/libnss_sss.so.2. Do you need "set solib-search-path" or "set sysroot"? warning: Could not load shared library symbols for SINFO: 0x. Do you need "set solib-search-path" or "set sysroot"? [Thread debugging using libthread_db enabled] Using host libthread_db library "/lib64/libthread_db.so.1". Core was generated by `/usr/sbin/glusterfsd -s bxts470192.eu.rabonet.com --volfile-id prod_xvavol.bxts'. Program terminated with signal 11, Segmentation fault. #0 glusterfs_handle_brick_status (req=0x7fae38001910) at glusterfsd-mgmt.c:1029 1029 any = active->first; (gdb) bt #0 glusterfs_handle_brick_status (req=0x7fae38001910) at glusterfsd-mgmt.c:1029 #1 0x00007fae4a5444f2 in synctask_wrap (old_task=<optimized out>) at syncop.c:375 #2 0x00007fae48b87d40 in ?? () from /lib64/libc.so.6 #3 0x0000000000000000 in ?? () Version-Release number of selected component (if applicable): 3.3 How reproducible: rarely on WA cluster Steps to Reproduce: 1. create a gluster volume 2. import to WA 3. make use of the volume. they crash after a certain time. Actual results: Crashes once in a while. Expected results: shouldn't crash. Additional info: Exact way to reproduce this is unknown and the details we have so far are, it looks like a race between get-state detail and profile command in various order. These are the commands the WA set up issues on gluster during the crash. There are times these commands work will in various orders. once in a while they crash. The RCA for this is tracked using the bug : https://bugzilla.redhat.com/show_bug.cgi?id=1596371 There is a similar bug to track the RCA : https://bugzilla.redhat.com/show_bug.cgi?id=1576726 Both end up in a situation which is not supposed to happen in the same environment.
REVIEW: https://review.gluster.org/20498 (core: dereference check on the variables in glusterfs_handle_brick_status) posted (#1) for review on master by hari gowtham
COMMIT: https://review.gluster.org/20498 committed in master by "Amar Tumballi" <amarts> with a commit message- core: dereference check on the variables in glusterfs_handle_brick_status problem: In a race condition, the active->first which is supposed to be filled is NULL and trying to dereference it crashs. back trace: Core was generated by `/usr/sbin/glusterfsd -s bxts470192.eu.rabonet.com --volfile-id prod_xvavol.bxts'. Program terminated with signal 11, Segmentation fault. 1029 any = active->first; (gdb) bt Change-Id: Ia6291865319a9456b8b01a5251be2679c4985b7c fixes: bz#1600451 Signed-off-by: Hari Gowtham <hgowtham>
This bug is getting closed because a release has been made available that should address the reported issue. In case the problem is still not fixed with glusterfs-5.0, please open a new bug report. glusterfs-5.0 has been announced on the Gluster mailinglists [1], packages for several distributions should become available in the near future. Keep an eye on the Gluster Users mailinglist [2] and the update infrastructure for your distribution. [1] https://lists.gluster.org/pipermail/announce/2018-October/000115.html [2] https://www.gluster.org/pipermail/gluster-users/