Bug 762012 (GLUSTER-280) - simple stripe, with write-behind set up, when dbench is run client crashes.
Summary: simple stripe, with write-behind set up, when dbench is run client crashes.
Keywords:
Status: CLOSED CURRENTRELEASE
Alias: GLUSTER-280
Product: GlusterFS
Classification: Community
Component: write-behind
Version: 2.0.6
Hardware: All
OS: Linux
low
low
Target Milestone: ---
Assignee: Raghavendra G
QA Contact:
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2009-09-24 05:16 UTC by Vinayak Hegde
Modified: 2009-09-24 08:35 UTC (History)
1 user (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed:
Regression: ---
Mount Type: ---
Documentation: ---
CRM:
Verified Versions:


Attachments (Terms of Use)

Description Vinayak Hegde 2009-09-24 02:42:12 UTC
without write-behind stripe runs well.

Comment 1 Vinayak Hegde 2009-09-24 05:16:06 UTC
simple stripe set-up.
performance translators on the client side are 
writ-behind and io-cache

when dbench is run on this set-up
it crashes with the following core.

(gdb) bt
#0  0x00002aaaac32cfd4 in __wb_collapse_write_bufs (
    requests=0x2aaab40a44f8, page_size=131072) at write-behind.c:1354
#1  0x00002aaaac32dc68 in wb_process_queue (frame=0x69a4f0,
    file=0x2aaab40a44c0, flush_all=0 '\0') at write-behind.c:1435
#2  0x00002aaaac32e0b3 in wb_ffr_cbk (frame=0x69a4f0,
    cookie=<value optimized out>, this=<value optimized out>, op_ret=0,
    op_errno=0) at write-behind.c:1694
#3  0x00002aaaac1258a8 in stripe_stack_unwind_cbk (frame=0x69a580,
    cookie=0x6c6ca0, this=0x60f4c0, op_ret=0, op_errno=0) at stripe.c:231
#4  0x00002aaaabf06f67 in client_flush_cbk (frame=0x6c6ca0,
    hdr=0x2aaab4044440, hdrlen=<value optimized out>,
    iobuf=<value optimized out>) at client-protocol.c:4748
#5  0x00002aaaabf067da in protocol_client_pollin (this=0x60e010,
    trans=0x612c70) at client-protocol.c:6344
#6  0x00002aaaabf0d3a2 in notify (this=0x2aaab40a44f8, event=2,
    data=0x612c70) at client-protocol.c:6387
#7  0x00002aaaacd90d13 in socket_event_handler (fd=<value optimized out>,
    idx=0, data=0x612c70, poll_in=1, poll_out=0, poll_err=0) at socket.c:814
#8  0x00002aaaaacee2c5 in event_dispatch_epoll (event_pool=0x609840)
    at event.c:804
#9  0x0000000000403e0e in main (argc=6, argv=0x7fff0a37a0b8)
    at glusterfsd.c:1325
(gdb) fr 0
#0  0x00002aaaac32cfd4 in __wb_collapse_write_bufs (
    requests=0x2aaab40a44f8, page_size=131072) at write-behind.c:1354
1354                    if ((request->stub->fop != GF_FOP_WRITE)
(gdb) p request->stub
$4 = (call_stub_t *) 0x0
(gdb)

Comment 2 Anand Avati 2009-09-24 05:18:45 UTC
PATCH: http://patches.gluster.com/patch/1484 in master (performance/write-behind: Add a NULL check for request->stub before checking request->stub->fop.)

Comment 3 Anand Avati 2009-09-24 05:19:19 UTC
PATCH: http://patches.gluster.com/patch/1483 in release-2.0 (performance/write-behind: Add a NULL check for request->stub before checking request->stub->fop.)


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