Hot-swapping disks

[Hauke Fath <hf%spg.tu-darmstadt.de@localhost>]

Hi,

what is the state with NetBSD and hot-swapping disk drives?

When I pulled a disk from a SAS expander backplane hanging off an LSI 
9207, I got

[...]
[ 1062.7357248] mpii0: physical device removed from slot 18
[ 1062.8357618] sd6: detached
[ 1110.4832140] panic: assert_sleepable: interrupt 
caller=0xffffffff809efe26
[ 1110.5284582] cpu0: Begin traceback...
[ 1110.5513370] vpanic() at netbsd:vpanic+0x160
[ 1110.5732316] snprintf() at netbsd:snprintf
[ 1110.6034409] assert_sleepable() at netbsd:assert_sleepable+0x9c
[ 1110.6367222] pool_cache_get_paddr() at 
netbsd:pool_cache_get_paddr+0x176
[ 1110.6832707] kmem_intr_alloc() at netbsd:kmem_intr_alloc+0x70
[ 1110.7143071] kern_malloc() at netbsd:kern_malloc+0x4d
[ 1110.7460265] mpii_event_sas() at netbsd:mpii_event_sas+0x91
[ 1110.7841959] mpii_intr() at netbsd:mpii_intr+0x20d
[ 1110.8140449] Xhandle_ioapic_level6() at 
netbsd:Xhandle_ioapic_level6+0x6a
[ 1110.8546571] --- interrupt ---
[ 1110.8738963] x86_stihlt() at netbsd:x86_stihlt+0x6
[ 1110.9035879] acpicpu_cstate_idle_enter() at 
netbsd:acpicpu_cstate_idle_enter+0xdb
[ 1110.9460720] acpicpu_cstate_idle() at netbsd:acpicpu_cstate_idle+0xba
[ 1110.9856992] idle_loop() at netbsd:idle_loop+0xcc
[ 1111.0156489] cpu0: End traceback...

[ 1111.0400905] dumping to dev 18,1 (offset=79728335, size=8384038):
[ 1111.0400905] dump device bad

-- this is a mirrored zfs setup on netbsd-9.

So what is the protocol for pulling a (SAS) disk from a zfs pool, when 
hot-swapping does not work?

Cheerio,
Hauke

-- 
     The ASCII Ribbon Campaign                    Hauke Fath
()     No HTML/RTF in email            Institut für Nachrichtentechnik
/\     No Word docs in email                     TU Darmstadt
     Respect for open standards              Ruf +49-6151-16-21344