Date: Fri, 5 Nov 2021 12:59:47 -0400 From: Chris Ross <cross+freebsd@distal.com> To: freebsd-fs <freebsd-fs@freebsd.org> Subject: Re: ZFS operations hanging, but no visible errors? Message-ID: <86999084-7007-4F08-A4C4-4A835A7E1C78@distal.com> In-Reply-To: <20211105173935.7aa53269@fabiankeil.de> References: <B28E52F4-F475-4CF6-BE0C-F5C803AD5757@distal.com> <20211105173935.7aa53269@fabiankeil.de>
next in thread | previous in thread | raw e-mail | index | archive | help
--Apple-Mail=_9D4F7EDC-2107-491F-8DD4-205632C97D87
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
charset=utf-8
> On Nov 5, 2021, at 12:39, Fabian Keil <freebsd-listen@fabiankeil.de> =
wrote:
> You can check with "procstat -kk" where the hanging
> processes are stuck.
Running prostate -kk on the rsync that was hung, then killed, then =
SIGKILL=E2=80=99d shows:
procstat -kk 35220
PID TID COMM TDNAME KSTACK
35220 102499 rsync - mi_switch+0xc1 =
_sleep+0x1cb vm_wait_doms+0xe2 vm_wait_domain+0x51 =
vm_domain_alloc_fail+0x86 vm_page_alloc_domain_after+0x7e =
uma_small_alloc+0x58 keg_alloc_slab+0xba zone_import+0xee =
zone_alloc_item+0x6f abd_alloc_chunks+0x61 abd_alloc+0x102 =
arc_hdr_alloc_abd+0xb0 arc_hdr_alloc+0x11e arc_read+0x4f4 =
dbuf_issue_final_prefetch+0x108 dbuf_prefetch_impl+0x3d0 =
dmu_zfetch+0x558
>=20
> The output of:
> dtrace -qn 'zfs-dbgmsg {printf("%Y: %s\n", walltimestamp, =
stringof(arg0))}'
> could be useful as well.
I=E2=80=99ve run=20
% sudo dtrace -qn 'zfs-dbgmsg {printf("%Y: %s\n", walltimestamp, =
stringof(arg0))}=E2=80=99
But, it=E2=80=99s yielding no output. Is that normal because it outputs =
information about operations (which aren=E2=80=99t happening), or should =
it=E2=80=99s output have been immediate?
- Chris=
--Apple-Mail=_9D4F7EDC-2107-491F-8DD4-205632C97D87--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?86999084-7007-4F08-A4C4-4A835A7E1C78>