[Gluster-devel] brick stop responding (3.4.0qa8)

[Emmanuel Dreyfus]

3.4.0qa8 works fine, but after a while, a NetBSD/amd64 brick stops 
responding (while NetBSD/i386 servers seems to work fine, therefore 
it looks 64 bits specific). ktrace shows it loops around poll calls. Here  
is what I can see if I stop it and inspect with gdb:

#0  0x00007f7ff62759da in ___lwp_park50 () from /lib/libc.so.12
(gdb) bt
#0  0x00007f7ff62759da in ___lwp_park50 () from /lib/libc.so.12
#1  0x00007f7ff6c086b9 in pthread_cond_timedwait ()
   from /usr/lib/libpthread.so.1
#2  0x00007f7ff200abd1 in iot_worker (data=0x7f7ff6fe3120) at io-threads.c:157
#3  0x00007f7ff6c09d75 in ?? () from /usr/lib/libpthread.so.1
#4  0x00007f7ff62759f0 in ___lwp_park50 () from /lib/libc.so.12
#5  0x00007f7fee000000 in ?? ()
#6  0x00007f7ff7fffcc0 in ?? ()
#7  0x0000000111110001 in ?? ()
#8  0x0000000033330003 in ?? ()
#9  0x0000000000000000 in ?? ()
(gdb) frame 2
#2  0x00007f7ff200abd1 in iot_worker (data=0x7f7ff6fe3120) at io-threads.c:157
157                                     ret = pthread_cond_timedwait (&conf->cond,
(gdb) list
152                                     pri = -1;
153                             }
154                             while (conf->queue_size == 0) {
155                                     conf->sleep_count++;
156     
157                                     ret = pthread_cond_timedwait (&conf->cond,
158                                                                   &conf->mutex,
159                                                                   &sleep_till);
160                                     conf->sleep_count--;
161     
(gdb) print conf->cond 
$1 = {ptc_magic = 1431633925, ptc_lock = 0 '\000', ptc_waiters = {
    ptqh_first = 0x7f7feec00000, ptqh_last = 0x7f7fee000230}, 
  ptc_mutex = 0x7f7ff6fe3120, ptc_private = 0x0}
(gdb) print conf->mutex
$2 = {ptm_magic = 858980355, ptm_errorcheck = 0 '\000', ptm_pad1 = "\000\000", 
  ptm_interlock = 0 '\000', ptm_pad2 = "\000\000", ptm_owner = 0x0, 
  ptm_waiters = 0x0, ptm_recursed = 0, ptm_spare2 = 0x0}

NB: ptc_magic and ptm_magic are correct.


(gdb) print sleep_till
$3 = {tv_sec = 1359902843, tv_nsec = 0}

This is also fine.


(gdb) info threads
  Id   Target Id         Frame 
* 1    LWP 1             0x00007f7ff200abd1 in iot_worker (data=0x7f7ff6fe3120)
    at io-threads.c:157

The current thread <Thread ID 1> has terminated.  See `help thread'.
(gdb) 

Is'nt it supposed to have multiple threads? 

-- 
Emmanuel Dreyfus
http://hcpnet.free.fr/pubz
manu at netbsd.org