[Bug 194513] zfs recv hangs in state kmem arena

[bugzilla-noreply at freebsd.org]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=194513

--- Comment #17 from Twingly <shared+bugs.freebsd.org at twingly.com> ---
We are also seeing this on our backup server
64 GB RAM
4x6TB (zpool with 2 mirrors)
*no* SSD ZIL or L2Arc

zfs send ... | ssh backupserver zfs receive zfs receive -u ...

often hangs, and I have to reboot the server to get rid of the hang process and
to be able to receive datasets again

no problem with 'zpool history', it runs fine


$ uname -a
FreeBSD hodor.live.lkp.primelabs.se 10.2-RELEASE-p18 FreeBSD 10.2-RELEASE-p18
#0: Sat May 28 08:53:43 UTC 2016    
root at amd64-builder.daemonology.net:/usr/obj/usr/src/sys/GENERIC  amd64

$ freebsd-version
10.2-RELEASE-p19

$ ps -l -U backup
 UID   PID PPID CPU PRI NI   VSZ  RSS MWCHAN   STAT TT     TIME COMMAND
1006 59745    1   0  20  0 42248 3164 kmem are Ds    -  0:03.81 zfs receive -u
storage/backup/mysql/leaf/data

$ procstat 59745
  PID  PPID  PGID   SID  TSID THR LOGIN    WCHAN     EMUL          COMM
59745     1 59745 59745     0   1 backup   kmem are  FreeBSD ELF64 zfs

$ sudo procstat -kk 59745
Password:
  PID    TID COMM             TDNAME           KSTACK
59745 100809 zfs              -                mi_switch+0xe1 sleepq_wait+0x3a
_cv_wait+0x16d vmem_xalloc+0x568 vmem_alloc+0x3d kmem_malloc+0x33
uma_large_malloc+0x49 malloc+0x43 dmu_recv_stream+0x114 zfs_ioc_recv+0x955
zfsdev_ioctl+0x5ca devfs_ioctl_f+0x139 kern_ioctl+0x255 sys_ioctl+0x140
amd64_syscall+0x357 Xfast_syscall+0xfb

$ sysctl -h hw.physmem  vm.kmem_size vm.kmem_size_max
hw.physmem: 68,578,963,456
vm.kmem_size: 66,774,925,312
vm.kmem_size_max: 1,319,413,950,874


I will look into raising vm.kmem_size, but right now I have to reboot hand
hopefully get some backups going...

-- 
You are receiving this mail because:
You are the assignee for the bug.