Avati, I've updated GlusterFS to patch-161 half an hour ago. The bug is still there. :( As I've found out, it's related with posix/locks, not io-threads. If posix/locks is turned off in server spec, then KDE starts flawlessly. Otherwise glusterfsd dies with the following backtrace log: -------- [May 07 15:33:04] [CRITICAL/common-utils.c:215/gf_print_trace()] debug-backtrace:Got signal (11), printing backtrace [May 07 15:33:04] [CRITICAL/common-utils.c:217/gf_print_trace()] debug-backtrace:/usr/lib64/libglusterfs.so.0(gf_print_trace+0x21) [0x2b25736faf71] [May 07 15:33:04] [CRITICAL/common-utils.c:217/gf_print_trace()] debug-backtrace:/lib64/libc.so.6 [0x2b25741725b0] [May 07 15:33:04] [CRITICAL/common-utils.c:217/gf_print_trace()] debug-backtrace:/usr/lib64/glusterfs/1.3.0-pre3/xlator/features/posix-locks.so [0x2aaaaacb47b7] [May 07 15:33:04] [CRITICAL/common-utils.c:217/gf_print_trace()] debug-backtrace:/usr/lib64/glusterfs/1.3.0-pre3/xlator/protocol/server.so [0x2aaaaaebf4fb] [May 07 15:33:04] [CRITICAL/common-utils.c:217/gf_print_trace()] debug-backtrace:/usr/lib64/glusterfs/1.3.0-pre3/xlator/protocol/server.so [0x2aaaaaeba694] [May 07 15:33:04] [CRITICAL/common-utils.c:217/gf_print_trace()] debug-backtrace:/usr/lib64/libglusterfs.so.0(sys_epoll_iteration+0xd4) [0x2b25736fc7e4] [May 07 15:33:04] [CRITICAL/common-utils.c:217/gf_print_trace()] debug-backtrace:[glusterfsd] [0x40164c] [May 07 15:33:04] [CRITICAL/common-utils.c:217/gf_print_trace()] debug-backtrace:/lib64/libc.so.6(__libc_start_main+0xf4) [0x2b257415fae4] [May 07 15:33:04] [CRITICAL/common-utils.c:217/gf_print_trace()] debug-backtrace:[glusterfsd] [0x401149] --------- My server.vol spec contains: -------- volume disk type storage/posix # POSIX FS translator option directory /mnt/hd # Export this directory end-volume #volume brick # type features/posix-locks # subvolumes disk #end-volume volume brick type performance/io-threads option thread-count 8 subvolumes disk end-volume ### Add network serving capability to above brick volume server type protocol/server option transport-type tcp/server # For TCP/IP transport # option bind-address 192.168.1.10 # Default is to listen on all interfaces # option listen-port 6996 # Default is 6996 option client-volume-filename /etc/glusterfs/client.vol subvolumes brick option auth.ip.brick.allow 10.1.0.* # Allow access to "brick" volume end-volume ------- With best regards, Andrey 2007/5/5, Anand Avati <avati@xxxxxxxxxxxxx>:
Andrey, can you please confirm the bug with the latest TLA checkout? few fixes have been committed in various places along with io-threads.
