[sheepdog-users] SIGABRT when doing: dog vdi check
Hitoshi Mitake
mitake.hitoshi at gmail.com
Tue Jan 7 14:19:23 CET 2014
At Tue, 07 Jan 2014 10:45:33 +0100,
Marcin Mirosław wrote:
>
> W dniu 07.01.2014 02:57, Hitoshi Mitake pisze:
> > At Mon, 06 Jan 2014 11:41:16 +0900,
> > Hitoshi Mitake wrote:
> >>
> >> At Sat, 4 Jan 2014 13:28:27 +0800,
> >> Liu Yuan wrote:
> >>>
> >>> On Fri, Jan 03, 2014 at 10:51:26PM +0100, Marcin Mirosław wrote:
> >>>> Hi!
> >>>> I'm new on "sheep-run";) I'm starting to try sheepdog so probably I'm
> >>>> doing many things wrongly.
> >>>> I'm playing with sheepdog-0.7.6.
> >>>>
> >>>> First problem (SIGABRT):
> >>>> I started multi sheep daemeon on localhost:
> >>>> # for x in 0 1 2 3 4; do sheep -c local -j size=128M -p 700$x
> >>>> /mnt/sheep/metadata/$x,/mnt/sheep/storage/$x; done
> >>>>
> >>>> Next:
> >>>> # dog cluster info
> >>>> Cluster status: Waiting for cluster to be formatted
> >>>>
> >>>> # dog cluster format -c 2:1
> >>>
> >>> 0.7.6 doesn't support erasure code. Try latest master branch
> >>
> >> Current stable-0.7 doesn't treat copy policies like "2:1" as an
> >> error. It would be confusing for users. I'll write a patch only for
> >> stable-0.7 for treating these options as errors.
> >>
> >>>
> >>>> using backend plain store
> >>>> # dog cluster info
> >>>> Cluster status: running, auto-recovery enabled
> >>>>
> >>>> Cluster created at Fri Jan 3 20:33:43 2014
> >>>>
> >>>> Epoch Time Version
> >>>> 2014-01-03 20:33:43 1 [127.0.0.1:7000, 127.0.0.1:7001,
> >>>> 127.0.0.1:7002, 127.0.0.1:7003, 127.0.0.1:7004]
> >>>> # dog vdi create testowy 5G
> >>>> # gdb -q dog
> >>>> Reading symbols from /usr/sbin/dog...Reading symbols from
> >>>> /usr/lib64/debug/usr/sbin/dog.debug...done.
> >>>> done.
> >>>> (gdb) set args vdi check testowy
> >>>> (gdb) run
> >>>> Starting program: /usr/sbin/dog vdi check testowy
> >>>> warning: Could not load shared library symbols for linux-vdso.so.1.
> >>>> Do you need "set solib-search-path" or "set sysroot"?
> >>>> warning: File "/lib64/libthread_db-1.0.so" auto-loading has been
> >>>> declined by your `auto-load safe-path' set to
> >>>> "$debugdir:$datadir/auto-load".
> >>>> To enable execution of this file add
> >>>> add-auto-load-safe-path /lib64/libthread_db-1.0.so
> >>>> line to your configuration file "/root/.gdbinit".
> >>>> To completely disable this security protection add
> >>>> set auto-load safe-path /
> >>>> line to your configuration file "/root/.gdbinit".
> >>>> For more information about this security protection see the
> >>>> "Auto-loading safe path" section in the GDB manual. E.g., run from the
> >>>> shell:
> >>>> info "(gdb)Auto-loading safe path"
> >>>> warning: Unable to find libthread_db matching inferior's thread library,
> >>>> thread debugging will not be available.
> >>>> PANIC: can't find next new idx
> >>>
> >>> seems that 0.7.x series is cracky about it. Hitoshi, can you verify
> >>> this?
> >>
> >> OK, I'll dig in the problem soon.
> >
> > Hi Marcin,
> >
> > I've fixed a serious bug which might produce the bug you've
> > suffered. The commit is this:
> > https://github.com/sheepdog/sheepdog/commit/b82632b47978315a50e9ba0bbad59f56453f63f5
> >
> > The patch is already backported to stable-0.7. Could you try the
> > latest stable-0.7? You can obtain the source code by the below step:
> >
> > git clone https://github.com/sheepdog/sheepdog.git
> > cd sheepdog
> > git checkout -b 0.7 origin/stable-0.7
> >
> > But the problem might depend on timing heavily. So reproducing the
> > problem will be not so easy... If you have time, I'd like you to try
> > the latest change.
>
> Hi all!
> Today I'm trying sheepdog on other computer and don;t have any problem
> to reproduce SIGABRT. I tried latest 0.7 stable (at 9de5329978) and "dog
> vdi check testowy" still crashes:
> # dog vdi check testowy
> PANIC: can't find next new idx
> dog exits unexpectedly (Aborted).
> dog() [0x40536a]
> /lib64/libpthread.so.0(+0xfd8f) [0x7f90aec31d8f]
> /lib64/libc.so.6(gsignal+0x38) [0x7f90ae8b2368]
> /lib64/libc.so.6(abort+0x147) [0x7f90ae8b36c7]
> dog() [0x40893a]
> dog() [0x40a3d9]
> dog() [0x40363e]
> /lib64/libc.so.6(__libc_start_main+0xf4) [0x7f90ae89ec04]
> dog() [0x4038a8]
> #7 0x000000000040faa5 in sd_backtrace () at logger.c:914
>
> addrs = {0x40f95d <sd_backtrace+29>, 0x40536b <crash_handler+43>,
> 0x7f90aec31d90 <__restore_rt>, 0x7f90ae8b2369 <__GI_raise+57>,
> 0x7f90ae8b36c8 <__GI_abort+328>, 0x40893b, 0x40a3da <vdi_check+186>,
> 0x40363f <main+1055>, 0x7f90ae89ec05 <__libc_start_main+2
> 45>, 0x4038a9 <_start+41>, 0x0 <repeats 1014 times>}
>
> i = <optimized out>
>
> n = <optimized out>
>
> __func__ = "sd_backtrace"
>
> #8 0x000000000040536b in crash_handler (signo=6) at dog.c:330
>
> __func__ = "crash_handler"
>
> #9 <signal handler called>
>
> No locals.
>
> #10 0x00007f90ae8b2369 in __GI_raise (sig=sig at entry=6) at
> ../nptl/sysdeps/unix/sysv/linux/raise.c:56
>
> resultvar = 0
>
> pid = 21938
>
> selftid = 21938
>
> #11 0x00007f90ae8b36c8 in __GI_abort () at abort.c:90
>
> save_stage = 2
>
> act = {__sigaction_handler = {sa_handler = 0x2, sa_sigaction = 0x2},
> sa_mask = {__val = {29029392, 102, 140259386245127, 5, 0, 208,
> 140259380182888, 0, 102, 320, 140259386270773, 0, 140259383961520, 0,
> 140259383961520, 140259383955552}}, sa_flags = -13587
> 68064, sa_restorer = 0x7f90ae9a3e80 <__memcpy_ssse3+9600>}
>
> sigs = {__val = {32, 0 <repeats 15 times>}}
>
> #12 0x000000000040893b in get_vnode_next_idx (nr_prev_idxs=<optimized
> out>, prev_idxs=<optimized out>, nr_entries=<optimized out>,
> entries=<optimized out>) at ../include/sheep.h:105
>
> i = <optimized out>
>
> idx = <optimized out>
>
> first_idx = <optimized out>
>
> found = <optimized out>
>
> #13 oid_to_vnodes (vnodes=0x7fffaefac8c0, nr_copies=2, oid=<optimized
> out>, nr_entries=320, entries=<optimized out>) at ../include/sheep.h:174
>
> idxs = {102, 0, 29029184, 0, 1, 0, 29029264, 0}
>
> i = <optimized out>
>
> vnodes = 0x7fffaefac8c0
>
> nr_copies = 2
>
> oid = <optimized out>
>
> nr_entries = 320
>
> #14 queue_vdi_check_work (oid=<optimized out>, done=done at entry=0x0,
> wq=wq at entry=0x1baf410, inode=0x7f90adc77010, inode=0x7f90adc77010) at
> vdi.c:1600
>
> info = 0x1baf580
>
> tgt_vnodes = {0xc2fd70 <sd_vnodes+5712>, 0x1baf3b8, 0x7f90adc77010,
> 0x414348 <create_worker_threads+104>, 0x7fffaefae207, 0x7f90adc76700,
> 0x1baf340, 0x1baf390}
>
> nr_copies = 2
>
> #15 0x000000000040a3da in vdi_check (argc=<optimized out>,
> argv=<optimized out>) at vdi.c:1634
>
> vdiname = 0x7fffaefae207 "testowy"
>
> ret = 0
>
> max_idx = <optimized out>
>
> done = 0
>
> vid = 13289526
>
> inode = 0x7f90adc77010
>
> wq = 0x1baf410
>
> __func__ = "vdi_check"
>
> #16 0x000000000040363f in main (argc=<optimized out>,
> argv=0x7fffaefacad8) at dog.c:494
>
> ch = <optimized out>
>
> longindex = 0
>
> ret = <optimized out>
>
> flags = 2
>
> long_options = 0xc2c560 <lopts.2390>
>
> commands = 0x1ba3010
>
> short_options = 0xc2e580 <sopts.2381"s:a:p:h"
>
> p = 0x41594d
> <__libc_csu_init+77"H\203\303\001H9\353u\352H\203\304\b[]A\\A]A^A_\303ff.\017\037\204"
>
> __func__ = "main"
>
>
> Additionally I'm getting QA Notice from portage:
> * QA Notice: Package triggers severe warnings which indicate that it
> * may exhibit random runtime failures.
> * group.c:348:45: warning: argument to ‘sizeof’ in ‘memcpy’ call is the
> same expression as the destination; did you mean to dereference it?
> [-Wsizeof-pointer-memaccess]
>
>
>
> Today it's gcc-4.8.2, but I tested with clang and it nothing changed.
> I'm configuring sheepdog using:
> # configure --prefix=/usr --build=x86_64-pc-linux-gnu
> --host=x86_64-pc-linux-gnu --mandir=/usr/share/man
> --infodir=/usr/share/info --datadir=/usr/share --sysconfdir=/etc
> --localstatedir=/var/lib --libdir=/usr/lib64 --disable-silent-rules
> --disable-dependency-tracking --docdir=/usr/share/doc/sheepdog-0.7.9999
> --disable-corosync --enable-sheepfs
>
> Dir "/mnt/sheep" is placed on one partition (I know this is slow but
> speed isn't important for me at this moment).
> If I can do something more to help please let me know.
Thanks a lot for your testing and detailed report, Marcin.
I backported an important patch for the above problem. Could you try
the latest stable-0.7?
BTW, it depends on sizeof(time_t). Could you try the below program on
your environment?
#include <stdio.h>
#include <time.h>
int main(void)
{
printf("sizeof(time_t): %lu\n", sizeof(time_t));
printf("sizeof(time_t *): %lu\n", sizeof(time_t *));
return 0;
}
If the sizes are different, the patch would fix the problem.
Thanks,
Hitoshi
More information about the sheepdog-users
mailing list