From: Hou Pu <houpu@bytedance.com>
To: Josef Bacik <josef@toxicpanda.com>, axboe@kernel.dk, mchristi@redhat.com
Cc: linux-block@vger.kernel.org, nbd@other.debian.org
Subject: Re: [PATCH] nbd: restore default timeout when setting it to zero
Date: Mon, 24 Aug 2020 11:23:18 +0800 [thread overview]
Message-ID: <701eaeb7-8b91-baa5-ebba-468f890c4cc5@bytedance.com> (raw)
In-Reply-To: <1accbf37-1a57-f072-7dc4-063fee991189@toxicpanda.com>
On 2020/8/21 9:57 PM, Josef Bacik wrote:
> On 8/21/20 3:21 AM, Hou Pu wrote:
>>
>>
>> On 2020/8/21 3:03 AM, Josef Bacik wrote:
>>> On 8/10/20 8:00 AM, Hou Pu wrote:
>>>> If we configured io timeout of nbd0 to 100s. Later after we
>>>> finished using it, we configured nbd0 again and set the io
>>>> timeout to 0. We expect it would timeout after 30 seconds
>>>> and keep retry. But in fact we could not change the timeout
>>>> when we set it to 0. the timeout is still the original 100s.
>>>>
>>>> So change the timeout to default 30s when we set it to zero.
>>>> It also behaves same as commit 2da22da57348 ("nbd: fix zero
>>>> cmd timeout handling v2").
>>>>
>>>> It becomes more important if we were reconfigure a nbd device
>>>> and the io timeout it set to zero. Because it could take 30s
>>>> to detect the new socket and thus io could be completed more
>>>> quickly compared to 100s.
>>>>
>>>> Signed-off-by: Hou Pu <houpu@bytedance.com>
>>>> ---
>>>> drivers/block/nbd.c | 2 ++
>>>> 1 file changed, 2 insertions(+)
>>>>
>>>> diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c
>>>> index ce7e9f223b20..bc9dc1f847e1 100644
>>>> --- a/drivers/block/nbd.c
>>>> +++ b/drivers/block/nbd.c
>>>> @@ -1360,6 +1360,8 @@ static void nbd_set_cmd_timeout(struct
>>>> nbd_device *nbd, u64 timeout)
>>>> nbd->tag_set.timeout = timeout * HZ;
>>>> if (timeout)
>>>> blk_queue_rq_timeout(nbd->disk->queue, timeout * HZ);
>>>> + else
>>>> + blk_queue_rq_timeout(nbd->disk->queue, 30 * HZ);
>>>> }
>>>> /* Must be called with config_lock held */
>>>>
>>>
>>> What about the tag_set.timeout? Thanks,
>>
>> I think user space could set io timeout to 0, thus we set
>> tag_set.timeout = 0 here and also we should tell the block layer
>> to restore 30s timeout in case it is not. tag_set.timeout == 0
>> imply 30s io timeout and retrying after timeout.
>>
>> (Sorry, I am not sure if I understand your question here. Could
>> you explain a little more if needed?)
>>
>
> I misunderstood what I was using the tagset timeout for. We don't want
> this here, if we're dropping a config for an nbd device and we want to
> reset it to defaults then we need to add this to nbd_config_put(). Thanks,
AFAIK If we killed a nbd server, then restarted it and reconfigured
the nbd socket, I think we might not reconfigure IO timeout to 0 since
nbd_config_put() is not called in such case. So could we still
restore default timeout here. Or am I missing something?
Thanks,
Hou
>
> Josef
next prev parent reply other threads:[~2020-08-24 3:23 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-08-10 12:00 [PATCH] nbd: restore default timeout when setting it to zero Hou Pu
2020-08-20 19:03 ` Josef Bacik
2020-08-21 7:21 ` Hou Pu
2020-08-21 13:57 ` Josef Bacik
2020-08-24 3:23 ` Hou Pu [this message]
2020-08-24 14:02 ` Josef Bacik
2020-08-25 8:27 ` Hou Pu
2020-08-25 17:29 ` Josef Bacik
2020-08-26 1:51 ` Hou Pu
2020-08-26 15:09 ` Jens Axboe
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=701eaeb7-8b91-baa5-ebba-468f890c4cc5@bytedance.com \
--to=houpu@bytedance.com \
--cc=axboe@kernel.dk \
--cc=josef@toxicpanda.com \
--cc=linux-block@vger.kernel.org \
--cc=mchristi@redhat.com \
--cc=nbd@other.debian.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox