[Bug 203906] ZFS lockup, spa_namespace_lock

[bugzilla-noreply at freebsd.org]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=203906

emz at norma.perm.ru changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |emz at norma.perm.ru

--- Comment #3 from emz at norma.perm.ru ---
I'm getting this on a recent r332062, after upgrade from 11.1-RELEASE-p6.
zfs [whatever] commands are freezing, totally:

# zfs list -t all
load: 9.48  cmd: zfs 12668 [spa_namespace_lock] 126.35r 0.00u 0.09s 0% 3912k
load: 7.44  cmd: zfs 12668 [spa_namespace_lock] 438.34r 0.00u 0.10s 0% 3236k

gstat is freezing (but after a minute is able to start showing data, however,
it contunues to freeze while showing data).

Since procstat is enormous, I'm attaching it as a file.

All the zfs pools are healthy, at least they were before the upgrade - now I
cannot get any output from zpool status.

This behavior is reproducible - onvce I was obvserving it on a less recent
-STABLE, but due to the lack of time I was forced to downgrade to 11.1-RELEASE.

ps:

[root at san1:~]# ps ax | grep D
  PID TT  STAT       TIME COMMAND
    0  -  DLs   462:29,65 [kernel]
    2  -  DL      0:00,00 [crypto]
    3  -  DL      0:00,00 [crypto returns]
    4  -  DL     30:55,60 [cam]
    5  -  DL      0:00,00 [soaiod1]
    6  -  DL      0:00,00 [soaiod2]
    7  -  DL      0:00,00 [soaiod3]
    8  -  DL      0:00,00 [soaiod4]
    9  -  DL     38:57,83 [zfskern]
   10  -  DL      0:00,00 [audit]
   13  -  DL      2:42,11 [geom]
   14  -  DL      0:02,46 [usb]
   15  -  DL      0:00,00 [sctp_iterator]
   16  -  DL      0:01,62 [pf purge]
   17  -  DL      0:05,83 [rand_harvestq]
   18  -  DL      0:00,01 [enc_daemon0]
   19  -  DL      0:00,01 [enc_daemon1]
   20  -  DL      0:00,11 [enc_daemon2]
   21  -  DL      0:00,60 [g_mirror swap]
   22  -  DL      1:04,21 [pagedaemon]
   23  -  DL      0:00,20 [vmdaemon]
   24  -  DNL     0:00,00 [pagezero]
   25  -  DL      0:00,04 [bufdaemon]
   26  -  DL      0:00,04 [bufspacedaemon]
   27  -  DL      0:00,13 [syncer]
   28  -  DL      0:00,07 [vnlru]
  740  -  DL      2:19,16 [ctl]
  898  -  DL      0:00,00 [ng_queue]
12174  -  D       0:00,75 /sbin/zfs list -t all
12236  -  D       0:00,78 /sbin/zfs list -t all
12240  -  D       0:00,77 /sbin/zfs list -t all
12338  -  D       0:00,75 /sbin/zfs list -t all
12340  -  D       0:00,75 /sbin/zfs list -t all
12364  -  DL      0:00,00 [ftcleanup]
12368  -  D       0:00,76 /sbin/zfs list -t all
12404  -  D       0:00,75 /sbin/zfs list -t all
12407  -  D       0:00,75 /sbin/zfs list -t all
12440  -  D       0:00,77 /sbin/zfs list -t all
12518  -  D       0:00,02 /usr/sbin/ctladm remove -b block -l 827
12520  -  D       0:00,09 /sbin/zfs list -t all
12554  -  D       0:00,09 /sbin/zfs list -t all
12639  -  D       0:00,08 /sbin/zfs list -t all
12651  -  D       0:00,09 /sbin/zfs list -t all
12716  -  D       0:00,09 zfs list -t snapshot
12721  -  D       0:00,09 /sbin/zfs list -t all
12741  -  D       0:00,00 /sbin/zfs destroy data/reference at ver13_2137
12743  -  D       0:00,00 /usr/sbin/ctladm remove -b block -l 809
12791  -  D       0:00,00 /sbin/zfs list -t all
12558  0  D+      0:00,31 gstat -do
12764  5  D+      0:00,00 gstat -d
12668  2  D+      0:00,09 zfs list -t all
 4898  3  D+      0:28,82 gstat
12814  6  S+      0:00,00 grep D
12737  4  D+      0:00,03 zfs destroy data/kvm/hv34/worker390
12747  8  D+      0:00,00 zfs get origin
data/kvm/desktop/desktop-master at desktop23

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