From: Rusty Russell <rusty@rustcorp.com.au>
To: Linux Kernel Developers List <linux-kernel@vger.kernel.org>
Cc: Frank Swiderski <fes@google.com>,
virtio-dev@lists.oasis-open.org, Theodore Ts'o <tytso@mit.edu>,
"Michael S. Tsirkin" <mst@redhat.com>,
virtualization@lists.linux-foundation.org
Subject: Re: [PATCH] virtio-blk: make the queue depth the max supportable by the hypervisor
Date: Mon, 17 Mar 2014 11:12:15 +1030 [thread overview]
Message-ID: <87siqhu048.fsf@rustcorp.com.au> (raw)
In-Reply-To: <1394854471-25700-1-git-send-email-tytso@mit.edu>
Theodore Ts'o <tytso@mit.edu> writes:
> The current virtio block sets a queue depth of 64, which is
> insufficient for very fast devices. It has been demonstrated that
> with a high IOPS device, using a queue depth of 256 can double the
> IOPS which can be sustained.
>
> As suggested by Venkatash Srinivas, set the queue depth by default to
> be one half the the device's virtqueue, which is the maximum queue
> depth that can be supported by the channel to the host OS (each I/O
> request requires at least two VQ entries).
>
> Also allow the queue depth to be something which can be set at module
> load time or via a kernel boot-time parameter, for
> testing/benchmarking purposes.
Note that with indirect descriptors (which is supported by Almost
Everyone), we can actually use the full index, so this value is a bit
pessimistic. But it's OK as a starting point.
Cheers,
Rusty.
WARNING: multiple messages have this Message-ID (diff)
From: Rusty Russell <rusty@rustcorp.com.au>
To: "Theodore Ts'o" <tytso@mit.edu>,
Linux Kernel Developers List <linux-kernel@vger.kernel.org>
Cc: "Theodore Ts'o" <tytso@mit.edu>,
Venkatesh Srinivas <venkateshs@google.com>,
"Michael S. Tsirkin" <mst@redhat.com>,
virtio-dev@lists.oasis-open.org,
virtualization@lists.linux-foundation.org,
Frank Swiderski <fes@google.com>
Subject: Re: [PATCH] virtio-blk: make the queue depth the max supportable by the hypervisor
Date: Mon, 17 Mar 2014 11:12:15 +1030 [thread overview]
Message-ID: <87siqhu048.fsf@rustcorp.com.au> (raw)
In-Reply-To: <1394854471-25700-1-git-send-email-tytso@mit.edu>
Theodore Ts'o <tytso@mit.edu> writes:
> The current virtio block sets a queue depth of 64, which is
> insufficient for very fast devices. It has been demonstrated that
> with a high IOPS device, using a queue depth of 256 can double the
> IOPS which can be sustained.
>
> As suggested by Venkatash Srinivas, set the queue depth by default to
> be one half the the device's virtqueue, which is the maximum queue
> depth that can be supported by the channel to the host OS (each I/O
> request requires at least two VQ entries).
>
> Also allow the queue depth to be something which can be set at module
> load time or via a kernel boot-time parameter, for
> testing/benchmarking purposes.
Note that with indirect descriptors (which is supported by Almost
Everyone), we can actually use the full index, so this value is a bit
pessimistic. But it's OK as a starting point.
Cheers,
Rusty.
next prev parent reply other threads:[~2014-03-17 0:42 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-03-14 23:57 [PATCH] virtio-blk: Initialize blkqueue depth from virtqueue size Venkatesh Srinivas
2014-03-15 3:34 ` [PATCH] virtio-blk: make the queue depth the max supportable by the hypervisor Theodore Ts'o
2014-03-15 3:34 ` Theodore Ts'o
2014-03-15 10:57 ` Konrad Rzeszutek Wilk
2014-03-15 10:57 ` Konrad Rzeszutek Wilk
2014-03-15 13:20 ` Theodore Ts'o
2014-03-15 13:20 ` Theodore Ts'o
2014-03-15 13:57 ` Christoph Hellwig
2014-03-15 13:57 ` Christoph Hellwig
2014-03-15 15:13 ` Theodore Ts'o
2014-03-15 15:13 ` Theodore Ts'o
2014-03-17 0:42 ` Rusty Russell [this message]
2014-03-17 0:42 ` Rusty Russell
2014-03-17 5:40 ` tytso
2014-03-17 5:40 ` tytso
2014-03-19 6:28 ` Rusty Russell
2014-03-19 6:28 ` Rusty Russell
2014-03-19 17:48 ` Venkatesh Srinivas
2014-03-19 17:48 ` Venkatesh Srinivas
2014-03-25 18:50 ` Venkatesh Srinivas
2014-03-25 18:50 ` Venkatesh Srinivas
2014-03-31 3:52 ` Rusty Russell
2014-03-31 3:52 ` Rusty Russell
2014-04-01 2:27 ` Theodore Ts'o
2014-04-01 10:49 ` Stefan Hajnoczi
2014-04-02 7:36 ` Rusty Russell
2014-04-01 2:27 ` Theodore Ts'o
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=87siqhu048.fsf@rustcorp.com.au \
--to=rusty@rustcorp.com.au \
--cc=fes@google.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mst@redhat.com \
--cc=tytso@mit.edu \
--cc=virtio-dev@lists.oasis-open.org \
--cc=virtualization@lists.linux-foundation.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.