lkml.org 
[lkml]   [2022]   [Aug]   [31]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
Subject[RFC PATCH V2 0/6] ublk_drv: add USER_RECOVERY support
Date
ublk_drv is a driver simply passes all blk-mq rqs to ublksrv[1] in
userspace. For each ublk queue, there is one ubq_daemon(pthread).
All ubq_daemons share the same process which opens /dev/ublkcX.
The ubq_daemon code infinitely loops on io_uring_enter() to
send/receive io_uring cmds which pass information of blk-mq
rqs.

Since the real IO handler(the process opening /dev/ublkcX) is in
userspace, it could crash if:
(1) the user kills -9 it because of IO hang on backend, system
reboot, etc...
(2) the process catches a exception(segfault, divisor error, oom...)
Therefore, the kernel driver has to deal with a dying process.

Now, if one ubq_daemon(pthread) or the process crashes, ublk_drv
must abort the dying ubq, stop the device and release everything.
This is not a good choice in practice because users do not expect
aborted requests, I/O errors and a released device. They may want
a recovery machenism so that no requests are aborted and no I/O
error occurs. Anyway, users just want everything works as uaual.

This RFC patchset implements USER_RECOVERY support. If the process
crashes, we allow ublksrv to provide new process and ubq_daemons.
We do not support single ubq_daemon(pthread) recovery because a
pthread rarely crashes.

Note: The responsibility of recovery belongs to the user who opens
/dev/ublkcX. After a process crash, the kernel driver only switch
the device's status to be ready for recovery or termination(STOP_DEV).
This patchset does not provide how to detect such a process crash in
userspace. A very straightfoward idea may be adding a watchdog.

Recovery feature is quite useful for real products. In detail,
we support this scenario:
(1) The /dev/ublkc0 is opened by process 0;
(2) Fio is running on /dev/ublkb0 exposed by ublk_drv and all
rqs are handled by process 0.
(3) Process 0 suddenly crashes(e.g. segfault);
(4) Fio is still running and submit IOs(but these IOs cannot
complete now)
(5) User recovers with process 1 and attach it to /dev/ublkc0
(6) All rqs are handled by process 1 now and IOs can be
completed now.

Note: The backend must tolerate double-write because we re-issue
a rq sent to the old(dying) process before. We allow users to
choose whether re-issue these rqs or not, please see patch 7 for
more detail.

We provide a sample script here to simulate the above steps:

***************************script***************************
LOOPS=10

__ublk_get_pid() {
pid=`./ublk list -n 0 | grep "pid" | awk '{print $7}'`
echo $pid
}

ublk_recover_kill()
{
for CNT in `seq $LOOPS`; do
dmesg -C
pid=`__ublk_get_pid`
echo -e "*** kill $pid now ***"
kill -9 $pid
sleep 6
echo -e "*** recover now ***"
./ublk recover -n 0
sleep 6
done
}

ublk_test()
{
dmesg -C
echo -e "*** add ublk device ***"
./ublk add -t null -d 4 -i 1
sleep 2
echo -e "*** start fio ***"
fio --bs=4k \
--filename=/dev/ublkb0 \
--runtime=140s \
--rw=read &
sleep 4
ublk_recover_kill
wait
echo -e "*** delete ublk device ***"
./ublk del -n 0
}

for CNT in `seq 4`; do
modprobe -rv ublk_drv
modprobe ublk_drv
echo -e "************ round $CNT ************"
ublk_test
sleep 5
done
***************************script***************************

You may run it with our modified ublksrv[2] which supports
recovey feature. No I/O error occurs and you can verify it
by typing
$ perf-tools/bin/tpoint block:block_rq_error

The basic idea of USER_RECOVERY is quite straightfoward:

(1) release/free everything belongs to the dying process.

Note: Since ublk_drv does save information about user process,
this work is important because we don't expect any resource
lekage. Particularly, ioucmds from the dying ubq_daemons
need to be completed(freed).

(2) init ublk queues including requeuing/aborting rqs.

(3) allow new ubq_daemons issue FETCH_REQ.

Here is steps to reocver:

(1) The monitor_work detects a crash, and it should requeue/abort inflight
rqs, complete old ioucmds and quiesce request queue to ban any incoming
ublk_queue_rq(). Then the ublk device is ready for a recovery/stop
procedure.

(2) For a user, after a process crash, he sends START_USER_RECOVERY
ctrl-cmd to /dev/ublk-control with a dev_id X (such as 3 for
/dev/ublkc3).

(2) Then ublk_drv should perpare for a new process to attach /dev/ublkcX.
All ublk_io structures are cleared and ubq_daemons are reset.

(3) Then, user should start a new process and ubq_daemons(pthreads) and
send FETCH_REQ by io_uring_enter() to make all ubqs be ready. The
user must correctly setup queues, flags and so on(how to persist
user's information is not related to this patchset).

(4) The user sends END_USER_RECOVERY ctrl-cmd to /dev/ublk-control with a
dev_id X.

(5) ublk_drv waits for all ubq_daemons getting ready. Then it unquiesces
request queue and new rqs are allowed.

You should use ublksrv[2] and tests[3] provided by us. We add 2 additional
tests to verify that recovery feature works. Our code will be PR-ed to
Ming's repo soon.

[1] https://github.com/ming1/ubdsrv
[2] https://github.com/old-memories/ubdsrv/tree/recovery-v1
[3] https://github.com/old-memories/ubdsrv/tree/recovery-v1/tests/generic

Since V1:
(1) refactor cover letter. Add intruduction on "how to detect a crash" and
"why we need recovery feature".
(2) do not refactor task_work and ublk_queue_rq().
(3) allow users freely stop/recover the device.
(4) add comment on ublk_cancel_queue().
(5) refactor monitor_work and aborting machenism since we add recovery
machenism in monitor_work.

ZiyangZhang (6):
ublk_drv: check 'current' instead of 'ubq_daemon'
ublk_drv: refactor ublk_cancel_queue()
ublk_drv: define macros for recovery feature and check them
ublk_drv: requeue rqs with recovery feature enabled
ublk_drv: consider recovery feature in aborting mechanism
ublk_drv: add START_USER_RECOVERY and END_USER_RECOVERY support

drivers/block/ublk_drv.c | 439 +++++++++++++++++++++++++++++++---
include/uapi/linux/ublk_cmd.h | 7 +
2 files changed, 419 insertions(+), 27 deletions(-)

--
2.27.0

\
 
 \ /
  Last update: 2022-08-31 17:53    [W:0.074 / U:0.084 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site