🚀 go-pugleaf

Witam.

Mam problem z FreeBSD 8.2-p4 64bit (kernel GENERIC) na maszynie z
Celeronem 220 1.2GHz, 2GB RAM.

Od kiedy zacz±³em u¿ywaæ ZFS, maszyna co kilka dni siê resetuje.
Load Average w momencie resetu raczej nie przekracza 0.1

Zajêto¶æ poola: 130GB z 1TB.
"zpool status" nie pokazuje b³êdów, "zpool scrub" te¿ ostatnio zrobiony.

Nie mam fizycznego dostêpu do maszyny, a wiêc nie bardzo mogê sprawdziæ
problemy sprzêtowe. Ale zanim zacz±³em u¿ywaæ ZFS, serwer by³ przez 3
miesi±ce stabilny.

Stacktrace pokazuje podobne miejsca w kodzie (fs/zfs/arc.c) przy ka¿dym resecie:

#v+
(...)
#7  0xffffffff80820f37 in vm_page_splay (pindex"3004, root=0x460700000c66)
    at /usr/src/sys/vm/vm_page.c:624
#8  0xffffffff80821932 in vm_page_remove (m=0xffffff007b206c70)
    at /usr/src/sys/vm/vm_page.c:741
#9  0xffffffff80821a7d in vm_page_free_toq (m=0xffffff007b206c70)
    at /usr/src/sys/vm/vm_page.c:1562
#10 0xffffffff8082085b in vm_object_page_remove (object=0xffffffff80b957a0, start"2976, end"3008, clean_only=0)
    at /usr/src/sys/vm/vm_object.c:1788
#11 0xffffffff80818c33 in vm_map_delete (map=0xffffff00010000e8, start=Variable "start" is not available.)
    at /usr/src/sys/vm/vm_map.c:2715
#12 0xffffffff80818d41 in vm_map_remove (map=0xffffff00010000e8, start446743524867047424, end446743524867178496)
    at /usr/src/sys/vm/vm_map.c:2846
#13 0xffffffff8080d6a5 in uma_large_free (slab=0xffffff00291c2470)
    at /usr/src/sys/vm/uma_core.c:3084
#14 0xffffffff805aff97 in free (addr=0xffffff8036700000, mtp=0xffffffff80f277c0)
    at /usr/src/sys/kern/kern_malloc.c:506
#15 0xffffffff80e36351 in arc_buf_destroy (buf=0xffffff002c2710d8, recycle=Variable "recycle" is not available.)
    at /usr/src/sys/modules/zfs/../../cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c:1497
#16 0xffffffff80e39614 in arc_evict (state=0xffffffff80f11b00, spa=0, bytes5238618, recycle=0, type=ARC_BUFC_DATA)
    at /usr/src/sys/modules/zfs/../../cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c:1780
#17 0xffffffff80e3a6ec in arc_adjust ()
    at /usr/src/sys/modules/zfs/../../cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c:1993
#18 0xffffffff80e3a9b0 in arc_reclaim_thread (dummy=Variable "dummy" is not available.)
    at /usr/src/sys/modules/zfs/../../cddl/contrib/opensolaris/uts/common/fs/zfs/arc.c:2251
#19 0xffffffff805994f8 in fork_exit ( callout=0xffffffff80e3a810 <arc_reclaim_thread>, arg=0x0, frame=0xffffff8092280c40)
    at /usr/src/sys/kern/kern_fork.c:845
#20 0xffffffff808954ae in fork_trampoline ()
    at /usr/src/sys/amd64/amd64/exception.S:565
#21 0x0000000000000000 in ?? ()
(...)
#v-

Pliki core.txt.* (z wyciêtymi procesami u¿ytkowników) na stronie:
http://kim.rolskiego.net/~simson/crash/

Dobrze szukam, podejrzewaj±c ZFS?

pozdrawiam,
Jacek