[Bugs] [Bug 1593826] Glusterfs being killed due to out of memory situation

bugzilla at redhat.com bugzilla at redhat.com
Sat Aug 11 19:47:46 UTC 2018


https://bugzilla.redhat.com/show_bug.cgi?id=1593826



--- Comment #31 from Amar Tumballi <atumball at redhat.com> ---
Tried to check the leak using valgrind on a 3.12 branch latest master...

==31275== 2,291,328 (313,344 direct, 1,977,984 indirect) bytes in 1,224 blocks
are definitely lost in loss record 524 of
 530
==31275==    at 0x4C2CB6B: malloc (vg_replace_malloc.c:299)
==31275==    by 0x4E888CB: mem_get (mem-pool.c:843)
==31275==    by 0x4E8897D: mem_get0 (mem-pool.c:741)
==31275==    by 0x12381B30: afr_fsync (afr-common.c:3607)
==31275==    by 0x1261201E: dht_fsync (dht-inode-read.c:998)
==31275==    by 0x12845F05: wb_fsync_helper (write-behind.c:1978)
==31275==    by 0x4E8499C: call_resume_keep_stub (call-stub.c:2582)
==31275==    by 0x1284B164: wb_do_winds (write-behind.c:1676)
==31275==    by 0x1284B27A: wb_process_queue (write-behind.c:1713)
==31275==    by 0x1284B336: wb_fulfill_cbk (write-behind.c:1058)
==31275==    by 0x12603DE5: dht_writev_cbk (dht-inode-write.c:119)
==31275==    by 0x1234DD20: afr_writev_unwind (afr-inode-write.c:246)
==31275== 
==31275== 5,263,980 (719,872 direct, 4,544,108 indirect) bytes in 2,812 blocks
are definitely lost in loss record 529 of
 530
==31275==    at 0x4C2CB6B: malloc (vg_replace_malloc.c:299)
==31275==    by 0x4E888CB: mem_get (mem-pool.c:843)
==31275==    by 0x4E8897D: mem_get0 (mem-pool.c:741)
==31275==    by 0x1235B93F: afr_changelog_do (afr-transaction.c:1685)
==31275==    by 0x1235C9DC: afr_changelog_post_op_now (afr-transaction.c:992)
==31275==    by 0x123815C5: afr_fsync_cbk (afr-common.c:3569)
==31275==    by 0x120ECFE1: client3_3_fsync_cbk (client-rpc-fops.c:954)
==31275==    by 0x513CE7F: rpc_clnt_handle_reply (rpc-clnt.c:778)
==31275==    by 0x513D17A: rpc_clnt_notify (rpc-clnt.c:971)
==31275==    by 0x5139742: rpc_transport_notify (rpc-transport.c:538)
==31275==    by 0x1103AB01: socket_event_poll_in (socket.c:2315)
==31275==    by 0x1103CF72: socket_event_handler (socket.c:2467)



------

The script used to reproduce the leak is:

python << EOF
import os, time

with open("/mnt/fuse/tmpfile", "w") as fh:
  while True:
    os.lseek(fh.fileno(), 0, os.SEEK_SET)
    os.write(fh.fileno(), str(time.time()).encode("ascii"))
    os.fsync(fh)
EOF

-- 
You are receiving this mail because:
You are on the CC list for the bug.
You are the assignee for the bug.


More information about the Bugs mailing list