From: Chuck Lever <chuck.lever@oracle.com>
To: Bruce Fields <bfields@fieldses.org>
Cc: Trond Myklebust <trondmy@hammerspace.com>,
Linux NFS Mailing List <linux-nfs@vger.kernel.org>,
Olga Kornievskaia <aglo@umich.edu>,
"bcodding@redhat.com" <bcodding@redhat.com>,
Jacob Shivers <jshivers@redhat.com>,
Frank Sorenson <fsorenso@redhat.com>
Subject: Re: unsharing tcp connections from different NFS mounts
Date: Wed, 20 Jan 2021 15:58:12 +0000 [thread overview]
Message-ID: <EED174E4-AA68-42AC-8099-AD4CAD29B441@oracle.com> (raw)
In-Reply-To: <20210119222229.GA29488@fieldses.org>
> On Jan 19, 2021, at 5:22 PM, bfields@fieldses.org wrote:
>
> On Wed, Oct 07, 2020 at 04:50:26PM +0000, Trond Myklebust wrote:
>> As far as I can tell, this thread started with a complaint that
>> performance suffers when we don't allow setups that hack the client by
>> pretending that a multi-homed server is actually multiple different
>> servers.
>>
>> AFAICS Tom Talpey's question is the relevant one. Why is there a
>> performance regression being seen by these setups when they share the
>> same connection? Is it really the connection, or is it the fact that
>> they all share the same fixed-slot session?
>>
>> I did see Igor's claim that there is a QoS issue (which afaics would
>> also affect NFSv3), but why do I care about QoS as a per-mountpoint
>> feature?
>
> Sorry for being slow to get back to this.
>
> Some more details:
>
> Say an NFS server exports /data1 and /data2.
>
> A client mounts both. Process 'large' starts creating 10G+ files in
> /data1, queuing up a lot of nfs WRITE rpc_tasks.
>
> Process 'small' creates a lot of small files in /data2, which requires a
> lot of synchronous rpc_tasks, each of which wait in line with the large
> WRITE tasks.
>
> The 'small' process makes painfully slow progress.
>
> The customer previously made things work for them by mounting two
> different server IP addresses, so the "small" and "large" processes
> effectively end up with their own queues.
>
> Frank Sorenson has a test showing the difference; see
>
> https://bugzilla.redhat.com/show_bug.cgi?id=1703850#c42
> https://bugzilla.redhat.com/show_bug.cgi?id=1703850#c43
>
> In that test, the "small" process creates files at a rate thousands of
> times slower when the "large" process is also running.
>
> Any suggestions?
Based on observation, there is a bottleneck in svc_recv which fully
serializes the receipt of RPC messages on a TCP socket. Large NFS
WRITE requests take longer to remove from the socket, and only one
nfsd can access that socket at a time.
Directing the large operations to a different socket means one nfsd
at a time can service those operations while other nfsd threads can
deal with the burst of small operations.
I don't know of any way to fully address this issue with a socket
transport other than by creating more transport sockets.
For RPC/RDMA I have some patches which enable svc_rdma_recvfrom()
to clear XPT_BUSY as soon as the ingress Receive buffer is dequeued.
--
Chuck Lever
next prev parent reply other threads:[~2021-01-20 16:00 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-10-06 15:13 unsharing tcp connections from different NFS mounts J. Bruce Fields
2020-10-06 15:20 ` Chuck Lever
2020-10-06 15:22 ` Bruce Fields
2020-10-06 17:07 ` Tom Talpey
2020-10-06 19:30 ` Bruce Fields
[not found] ` <CAGrwUG5_KeRVR8chcA8=3FSeii2+4c8FbuE=CSGAtYVYqV4kLg@mail.gmail.com>
2020-10-07 14:08 ` Tom Talpey
2020-10-06 19:36 ` Benjamin Coddington
2020-10-06 21:46 ` Olga Kornievskaia
2020-10-07 0:18 ` J. Bruce Fields
2020-10-07 11:27 ` Benjamin Coddington
2020-10-07 12:55 ` Benjamin Coddington
2020-10-07 13:45 ` Chuck Lever
2020-10-07 14:05 ` Bruce Fields
2020-10-07 14:15 ` Chuck Lever
2020-10-07 16:05 ` Bruce Fields
2020-10-07 16:44 ` Trond Myklebust
2020-10-07 17:15 ` Bruce Fields
2020-10-07 17:29 ` Trond Myklebust
2020-10-07 18:05 ` bfields
2020-10-07 19:11 ` Trond Myklebust
2020-10-07 20:29 ` bfields
2020-10-07 18:04 ` Benjamin Coddington
2020-10-07 18:19 ` Trond Myklebust
2020-10-07 16:50 ` Trond Myklebust
2021-01-19 22:22 ` bfields
2021-01-19 23:09 ` Trond Myklebust
2021-01-20 15:07 ` bfields
2021-05-03 20:09 ` bfields
2021-05-04 2:08 ` NeilBrown
2021-05-04 13:27 ` Tom Talpey
2021-05-04 14:27 ` Trond Myklebust
2021-05-04 16:51 ` bfields
2021-05-04 21:32 ` Daire Byrne
2021-05-04 21:48 ` Trond Myklebust
2021-05-05 12:53 ` Daire Byrne
2021-01-20 15:58 ` Chuck Lever [this message]
2020-10-07 13:56 ` Patrick Goetz
2020-10-07 16:28 ` Igor Ostrovsky
2020-10-07 16:30 ` Benjamin Coddington
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=EED174E4-AA68-42AC-8099-AD4CAD29B441@oracle.com \
--to=chuck.lever@oracle.com \
--cc=aglo@umich.edu \
--cc=bcodding@redhat.com \
--cc=bfields@fieldses.org \
--cc=fsorenso@redhat.com \
--cc=jshivers@redhat.com \
--cc=linux-nfs@vger.kernel.org \
--cc=trondmy@hammerspace.com \
/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;
as well as URLs for NNTP newsgroup(s).