Description of problem: In multinode slave setup, glusterd crashes on slave node other than the one which is used to create geo-rep sessoin during geo-rep create push-pem. Version-Release number of selected component (if applicable): mainline master branch. How reproducible: Always Steps to Reproduce: 1. Create master and slave volume, with slave volume spread across more than one node. 2. On master: gluster vol geo-rep gsec_create 3. On master: gluster vol geo-rep crate mastervol slave_host:slavevol create push-pem Actual results: glusterd crashed on the slave node other than the one which is used to create goe-rep session. Expected results: glusterd should not crash. Additional info:
REVIEW: http://review.gluster.org/9018 (glusterd/geo-rep: Fix glusterd crash in non-originator slave node.) posted (#1) for review on master by Kotresh HR (khiremat)
Stack trace of glusterd crash is follows: #0 0x00007f8cf479ac39 in raise () from /lib64/libc.so.6 #1 0x00007f8cf479c348 in abort () from /lib64/libc.so.6 #2 0x00007f8cf4793b96 in __assert_fail_base () from /lib64/libc.so.6 #3 0x00007f8cf4793c42 in __assert_fail () from /lib64/libc.so.6 #4 0x00007f8cf59b0107 in __gf_free (free_ptr=0x7f8cd40046c0) at mem-pool.c:251 #5 0x00007f8ceb288af7 in glusterd_op_copy_file (dict=dict@entry=0x7f8cd40010dc, op_errstr=op_errstr@entry=0xf0a5f8) at glusterd-geo-rep.c:4444 #6 0x00007f8ceb236e3b in glusterd_op_commit_perform (op=GD_OP_COPY_FILE, dict=dict@entry=0x7f8cd40010dc, op_errstr=op_errstr@entry=0xf0a5f8, rsp_dict=rsp_dict@entry=0x7f8cd4000f6c) at glusterd-op-sm.c:4940 #7 0x00007f8ceb23d25c in glusterd_op_ac_commit_op (event=0x7f8cd4000bc0, ctx=0x7f8cd4002850) at glusterd-op-sm.c:4640 #8 0x00007f8ceb23b649 in glusterd_op_sm () at glusterd-op-sm.c:6575 #9 0x00007f8ceb229456 in __glusterd_handle_commit_op (req=req@entry=0xd06d7c) at glusterd-handler.c:943 #10 0x00007f8ceb2280b0 in glusterd_big_locked_handler (req=0xd06d7c, actor_fn=0x7f8ceb229350 <__glusterd_handle_commit_op>) at glusterd-handler.c:79 #11 0x00007f8cf59bd342 in synctask_wrap (old_task=<optimized out>) at syncop.c:335 #12 0x00007f8cf47aca00 in ?? () from /lib64/libc.so.6 #13 0x0000000000000000 in ?? ()
COMMIT: http://review.gluster.org/9018 committed in master by Krishnan Parthasarathi (kparthas) ------ commit adfccb57565ff1af0bbb79efc3cf489c55509aba Author: Kotresh HR <khiremat> Date: Fri Oct 31 07:06:52 2014 +0530 glusterd/geo-rep: Fix glusterd crash in non-originator slave node. Problem: glusterd crashes in non-originator slave node during geo-rep create push-pem. Cause: In glusterd_op_copy_file, the value of the key "common_pem_contents" is freed explicitly even after dict_set is successful when it is taken cared by dict_free. Solution: Free only in failure cases before dict_set. Change-Id: I65b5f32ee2b946107ad279b1fe3d728ec699bc7e BUG: 1159119 Signed-off-by: Kotresh HR <khiremat> Reviewed-on: http://review.gluster.org/9018 Tested-by: Gluster Build System <jenkins.com> Reviewed-by: Poornima G <pgurusid> Reviewed-by: Krishnan Parthasarathi <kparthas> Tested-by: Krishnan Parthasarathi <kparthas>
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-3.7.0, please open a new bug report. glusterfs-3.7.0 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] http://thread.gmane.org/gmane.comp.file-systems.gluster.devel/10939 [2] http://thread.gmane.org/gmane.comp.file-systems.gluster.user