Hi all, Glusterfs 2.0.6 works on our site, but i see a lot of messages like this: [2009-10-29 11:42:04] E [server-protocol.c:6388:server_setdents] server: fd - 1 (19488922): received a null buffer, returning EINVAL and sometimes also a random crash of servers carrying the replicated namespace for unified volume, probably related to a simultaneous access to the same file from different nodes (not sure, but it seems to be the reason). The first thing i wanted to try out, was the latest release of glusterfs, but the server crashes immediately after start. Here is the content of log file: pending frames: patchset: v2.0.7 signal received: 11 time of crash: 2009-10-29 11:07:54 configuration details: argp 1 backtrace 1 dlfcn 1 fdatasync 1 libpthread 1 llistxattr 1 setfsid 1 spinlock 1 epoll.h 1 xattr.h 1 st_atim.tv_nsec 1 package-string: glusterfs 2.0.7 //lib/libpthread.so.0[0xf76b46aa] [0xffffe400] //lib/libpthread.so.0(sem_timedwait+0x6e)[0xf76b293d] ================================================================================ Version : glusterfs 2.0.7 built on Oct 29 2009 10:31:25 git: v2.0.7 Starting Time: 2009-10-29 11:07:54 Command line : glusterfsd -f /etc/glusterfs/server.vol PID : 25929 System name : Linux Nodename : prkom13 Kernel Release : 2.6.31.4 Hardware Identifier: x86_64 glusterfsd(gf_daemon+0xc2)[0x804aade] glusterfsd(main+0xe9a)[0x804babc] //lib/libc.so.6(__libc_start_main+0xb4)[0xf75d4500] glusterfsd[0x8049d49] --------- Given volfile: +------------------------------------------------------------------------------+ 1: volume scratchdir 2: type storage/posix 3: option directory /export/scratch 4: end-volume 5: 6: volume home-raw 7: type storage/posix 8: option directory /export/home 9: end-volume 10: 11: volume home-locks 12: type features/posix-locks 13: subvolumes home-raw 14: end-volume 15: 16: volume home-ns-raw 17: type storage/posix 18: option directory /export/home-ns 19: end-volume 20: 21: volume home-ns-locks 22: type features/posix-locks 23: subvolumes home-ns-raw 24: end-volume 25: 26: volume server 27: type protocol/server 28: option transport-type tcp/server 29: subvolumes home-locks home-ns-locks scratchdir 30: option transport.socket.listen-port 6996 31: option auth.addr.home-locks.allow 192.168.7.* 32: option auth.addr.home-ns-locks.allow 192.168.7.* 33: option auth.addr.scratchdir.allow 192.168.7.* 34: end-volume +------------------------------------------------------------------------------+ [2009-10-29 11:07:54] N [glusterfsd.c:1315:main] glusterfs: Successfully started pending frames: patchset: v2.0.7 signal received: 11 time of crash: 2009-10-29 11:07:54 configuration details: argp 1 backtrace 1 dlfcn 1 fdatasync 1 libpthread 1 llistxattr 1 setfsid 1 spinlock 1 epoll.h 1 xattr.h 1 st_atim.tv_nsec 1 package-string: glusterfs 2.0.7 //lib/libpthread.so.0[0xf76b46aa] [0xffffe400] //lib/libpthread.so.0(sem_post+0x7b)[0xf76b2758] glusterfsd(main+0x11f3)[0x804be15] //lib/libc.so.6(__libc_start_main+0xb4)[0xf75d4500] glusterfsd[0x8049d49] --------- My first question is, does it appear on other sites too, or am i alone with this problem? The installation runs on linux, kernel is 2.6.31.4, libfuse 2.8.1. Krzysztof
