ZFS panic: [Re: stable/10 panic under disk load]

Steven Hartland steven at multiplay.co.uk
Fri Nov 21 13:44:40 UTC 2014 

Not had chance to look at this I'm afraid, as I'm out at an event, but
we're actually using releng/10.1 here as a cache box which is doing a
between 2-8Gbps from a ZFS array so pretty high loads, with zero problems.

That and not other reports of issues I'm wondering if you have either some
bad data on your pool or a HW issue e.g. bad memory / processor?



On 21 November 2014 10:28, Dmitry Morozovsky <marck at rinet.ru> wrote:

> Steven, colleagues,
>
> any news on this? I now have a bunch of cores, most from my own
> experiments,
> but also from daily find, all with similar sympthoms.
>
> On Tue, 18 Nov 2014, Dmitry Morozovsky wrote:
>
> > On Tue, 18 Nov 2014, Dmitry Morozovsky wrote:
> >
> > > On Tue, 18 Nov 2014, Steven Hartland wrote:
> > >
> > > > Can u plug a usb drive in to get a dump?
> > >
> > > Hm, will it work over USB stack? I can try this.
> > >
> > > BTW: it seems some internal ZFS locking trouble exists, as trere are 3
> cases:
> > >
> > > pool/R/fs1 mounted as /fs1
> > > pool/R/fs2
> > > pool/R/fs3
> > >
> > > tar cf - /fs1 >/dev/null works ok
> > > tar cf - /fs2 >/dev/null works ok
> > > rsync -avHP /fs1/ /fs2/ panics in few minutes
> > >
> > > will try to configure dump to USB SATA
> >
> > wow, it works ;)
> >
> > not on the first trial, but anyway, here we go:
> >
> > #0  doadump (textdump=1621911824) at pcpu.h:219
> > #1  0xffffffff803471d5 in db_fncall (dummy1=<value optimized out>,
> > dummy2=<value optimized out>, dummy3=<value optimized out>,
> >     dummy4=<value optimized out>) at /usr/src/sys/ddb/db_command.c:568
> > #2  0xffffffff80346ebd in db_command (cmd_table=0x0) at
> > /usr/src/sys/ddb/db_command.c:440
> > #3  0xffffffff80346c34 in db_command_loop () at
> > /usr/src/sys/ddb/db_command.c:493
> > #4  0xffffffff80349580 in db_trap (type=<value optimized out>, code=0) at
> > /usr/src/sys/ddb/db_main.c:231
> > #5  0xffffffff80940cd9 in kdb_trap (type=3, code=0, tf=<value optimized
> out>)
> > at /usr/src/sys/kern/subr_kdb.c:656
> > #6  0xffffffff80ce8ca3 in trap (frame=0xfffffe0860ac6d40) at
> > /usr/src/sys/amd64/amd64/trap.c:556
> > #7  0xffffffff80ccf492 in calltrap () at
> > /usr/src/sys/amd64/amd64/exception.S:232
> > #8  0xffffffff8094043e in kdb_enter (why=0xffffffff80f5b27c "panic",
> msg=<value
> > optimized out>) at cpufunc.h:63
> > #9  0xffffffff80908f76 in vpanic (fmt=<value optimized out>, ap=<value
> > optimized out>) at /usr/src/sys/kern/kern_shutdown.c:752
> > #10 0xffffffff80908fe3 in panic (fmt=0xffffffff8154f850 "\004") at
> > /usr/src/sys/kern/kern_shutdown.c:688
> > #11 0xffffffff80b64502 in vm_fault_hold (map=<value optimized out>,
> > vaddr=<value optimized out>, fault_type=<value optimized out>,
> >     fault_flags=<value optimized out>, m_hold=<value optimized out>) at
> > /usr/src/sys/vm/vm_fault.c:341
> > #12 0xffffffff80b62b87 in vm_fault (map=0xfffff80002000000, vaddr=<value
> > optimized out>, fault_type=1 '\001', fault_flags=128)
> >     at /usr/src/sys/vm/vm_fault.c:281
> > #13 0xffffffff80ce9551 in trap_pfault (frame=0xfffffe0860ac7400,
> usermode=0) at
> > /usr/src/sys/amd64/amd64/trap.c:752
> > #14 0xffffffff80ce8cba in trap (frame=0xfffffe0860ac7400) at
> > /usr/src/sys/amd64/amd64/trap.c:440
> > #15 0xffffffff80ccf492 in calltrap () at
> > /usr/src/sys/amd64/amd64/exception.S:232
> > #16 0xffffffff81b69d04 in zap_leaf_lookup_closest (l=0xfffff801bd6ec880,
> > h=1441072784640835584, cd=1, zeh=0xfffffe0860ac7518)
> >     at
> >
> /usr/src/sys/modules/zfs/../../cddl/contrib/opensolaris/uts/common/fs/zfs/zap_leaf.c:466
> > #17 0xffffffff81b688ee in fzap_cursor_retrieve (zap=0xfffff8001676ce80,
> > zc=0xfffffe0860ac77d8, za=0xfffffe0860ac76c0)
> >     at
> >
> /usr/src/sys/modules/zfs/../../cddl/contrib/opensolaris/uts/common/fs/zfs/zap.c:1190
> > #18 0xffffffff81b6dc97 in zap_cursor_retrieve (zc=0xfffffe0860ac77d8,
> > za=0xfffffe0860ac76c0)
> >     at
> >
> /usr/src/sys/modules/zfs/../../cddl/contrib/opensolaris/uts/common/fs/zfs/zap_micro.c:1290
> > #19 0xffffffff81ba8f16 in zfs_freebsd_readdir (ap=<value optimized out>)
> >     at
> >
> /usr/src/sys/modules/zfs/../../cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_vnops.c:2565
> > #20 0xffffffff80e03967 in VOP_READDIR_APV (vop=<value optimized out>,
> a=<value
> > optimized out>) at vnode_if.c:1821
> > #21 0xffffffff809b1d1c in kern_getdirentries (td=0xfffff80025ff1490,
> fd=<value
> > optimized out>,
> >     buf=0x801428000 <Address 0x801428000 out of bounds>, count=<value
> optimized
> > out>, basep=0xfffffe0860ac7990, residp=0x0)
> >     at vnode_if.h:758
> > #22 0xffffffff809b1ad8 in sys_getdirentries (td=0xfffff801bd6ec880,
> > uap=0xfffffe0860ac7a40) at /usr/src/sys/kern/vfs_syscalls.c:4030
> > #23 0xffffffff80ce9aca in amd64_syscall (td=0xfffff80025ff1490,
> traced=0) at
> > subr_syscall.c:134
> > #24 0xffffffff80ccf77b in Xfast_syscall () at
> > /usr/src/sys/amd64/amd64/exception.S:391
> > #25 0x000000080091043a in ?? ()
> >
> > #16 0xffffffff81b69d04 in zap_leaf_lookup_closest (l=0xfffff801bd6ec880,
> > h=1441072784640835584, cd=1, zeh=0xfffffe0860ac7518)    at
> >
> /usr/src/sys/modules/zfs/../../cddl/contrib/opensolaris/uts/common/fs/zfs/zap_leaf.c:466
> > 466                             if (HCD_GTEQ(le->le_hash, le->le_cd, h,
> cd) &&
> > (kgdb) p *le->le_hash
> > No symbol "le" in current context.
> > (kgdb) p le
> > No symbol "le" in current context.
> > (kgdb) p h
> > $1 = 1441072784640835584
> > (kgdb) p *h
> > Cannot access memory at address 0x13ffb81000000000
> > (kgdb) p cd
> > $2 = 1
> >
> > where now?
> >
> > >
> > > >
> > > >
> > > > > On 18 Nov 2014, at 16:57, Dmitry Morozovsky <marck at rinet.ru>
> wrote:
> > > > >
> > > > >> On Tue, 18 Nov 2014, Dmitry Morozovsky wrote:
> > > > >>
> > > > >> my backup server after updrade to frest stable/10
> > > > >>
> > > > >> start panicing on heavy disk load like rsync at
> > > > >
> > > > > Yes, it is reproducible easy and now I'm at ddb prompt with
> > > > >
> > > > > cpuid = 0
> > > > > KDB: stack backtrace:
> > > > > db_trace_self_wrapper() at db_trace_self_wrapper+0x2b/frame
> 0xfffffe0860864d60
> > > > > kdb_backtrace() at kdb_backtrace+0x39/frame 0xfffffe0860864e10
> > > > > vpanic() at vpanic+0x126/frame 0xfffffe0860864e50
> > > > > panic() at panic+0x43/frame 0xfffffe0860864eb0
> > > > > vm_fault_hold() at vm_fault_hold+0x1932/frame 0xfffffe0860865100
> > > > > vm_fault() at vm_fault+0x77/frame 0xfffffe0860865140
> > > > > trap_pfault() at trap_pfault+0x201/frame 0xfffffe08608651e0
> > > > > trap() at trap+0x47a/frame 0xfffffe08608653f0
> > > > > calltrap() at calltrap+0x8/frame 0xfffffe08608653f0
> > > > > --- trap 0xc, rip = 0xffffffff81b69d04, rsp = 0xfffffe08608654b0,
> rbp =
> > > > > 0xfffffe0860865500 ---
> > > > > zap_leaf_lookup_closest() at zap_leaf_lookup_closest+0xb4/frame
> > > > > 0xfffffe0860865500
> > > > > fzap_cursor_retrieve() at fzap_cursor_retrieve+0x16e/frame
> 0xfffffe0860865570
> > > > > zap_cursor_retrieve() at zap_cursor_retrieve+0x1f7/frame
> 0xfffffe0860865600
> > > > > zfs_freebsd_readdir() at zfs_freebsd_readdir+0x426/frame
> 0xfffffe0860865840
> > > > > VOP_READDIR_APV() at VOP_READDIR_APV+0xa7/frame 0xfffffe0860865870
> > > > > kern_getdirentries() at kern_getdirentries+0x21c/frame
> 0xfffffe0860865970
> > > > > sys_getdirentries() at sys_getdirentries+0x28/frame
> 0xfffffe08608659a0
> > > > > amd64_syscall() at amd64_syscall+0x25a/frame 0xfffffe0860865ab0
> > > > > Xfast_syscall() at Xfast_syscall+0xfb/frame 0xfffffe0860865ab0
> > > > > --- syscall (196, FreeBSD ELF64, sys_getdirentries), rip =
> 0x80091043a, rsp =
> > > > > 0x7fffffffb538, rbp = 0x7fffffffb560 ---
> > > > > KDB: enter: panic
> > > > > [ thread pid 1167 tid 100461 ]
> > > > > Stopped at      kdb_enter+0x3e: movq    $0,kdb_why
> > > > > db>
> > > > >
> > > > > Can I obtain somthing useful from here?  I'm afraid it's not easy
> to attach
> > > > > additional disk for crash dumps to this server...
> > > > >
> > > > >
> > > > >>
> > > > >> FreeBSD whale.rinet.ru 10.1-STABLE FreeBSD 10.1-STABLE #195
> r274646: Tue Nov 18
> > > > >> 12:15:24 MSK 2014
> > > > >> marck at castor.rinet.ru:/usr/obj/FreeBSD/pristine/src.10/sys/GENERIC
> amd64
> > > > >>
> > > > >>
> > > > >> panic: vm_fault: fault on nofault entry, addr: fffffe001805b000
> > > > >> cpuid = 0
> > > > >> KDB: stack backtrace:
> > > > >> #0 0xffffffff80964fa0 at kdb_backtrace+0x60
> > > > >> #1 0xffffffff8092a085 at panic+0x155
> > > > >> #2 0xffffffff80ba168e at vm_fault_hold+0x1b6e
> > > > >> #3 0xffffffff80b9fad7 at vm_fault+0x77
> > > > >> #4 0xffffffff80d2861c at trap_pfault+0x19c
> > > > >> #5 0xffffffff80d27dea at trap+0x47a
> > > > >> #6 0xffffffff80d0db92 at calltrap+0x8
> > > > >> #7 0xffffffff819df8ee at fzap_cursor_retrieve+0x16e
> > > > >> #8 0xffffffff819e4c97 at zap_cursor_retrieve+0x1f7
> > > > >> #9 0xffffffff81a1fed6 at zfs_freebsd_readdir+0x426
> > > > >> #10 0xffffffff80e456b7 at VOP_READDIR_APV+0xa7
> > > > >> #11 0xffffffff809d68cc at kern_getdirentries+0x21c
> > > > >> #12 0xffffffff809d6688 at sys_getdirentries+0x28
> > > > >> #13 0xffffffff80d28da1 at amd64_syscall+0x351
> > > > >> #14 0xffffffff80d0de7b at Xfast_syscall+0xfb
> > > > >> Uptime: 1m51s
> > > > >>
> > > > >> Unfortunately it's ZFS only, so I have no space to white panic
> dump.
> > > > >>
> > > > >> I'm now trying to rebuild kernel with debugger turned on, as
> luckily I have
> > > > >> working console at SOL...
> > > > >>
> > > > >> Any preliminary hints?
> > > > >>
> > > > >>
> > > > >
> > > > > --
> > > > > Sincerely,
> > > > > D.Marck                                     [DM5020, MCK-RIPE,
> DM3-RIPN]
> > > > > [ FreeBSD committer:
>  marck at FreeBSD.org ]
> > > > >
> ------------------------------------------------------------------------
> > > > > *** Dmitry Morozovsky --- D.Marck --- Wild Woozle ---
> marck at rinet.ru ***
> > > > >
> ------------------------------------------------------------------------
> > > > _______________________________________________
> > > > freebsd-stable at freebsd.org mailing list
> > > > http://lists.freebsd.org/mailman/listinfo/freebsd-stable
> > > > To unsubscribe, send any mail to "
> freebsd-stable-unsubscribe at freebsd.org"
> > > >
> > >
> > >
> >
> >
>
> --
> Sincerely,
> D.Marck                                     [DM5020, MCK-RIPE, DM3-RIPN]
> [ FreeBSD committer:                                 marck at FreeBSD.org ]
> ------------------------------------------------------------------------
> *** Dmitry Morozovsky --- D.Marck --- Wild Woozle --- marck at rinet.ru ***
> ------------------------------------------------------------------------
>

More information about the freebsd-stable mailing list