Comments on: I Have a Schedule to Keep – IO Schedulers http://www.linux-mag.com/id/7564/ Open Source, Open Standards Sat, 05 Oct 2013 13:48:18 +0000 hourly 1 http://wordpress.org/?v=3.1 By: speed145a http://www.linux-mag.com/id/7564/#comment-7120 speed145a Wed, 30 Nov -0001 00:00:00 +0000 http://www.linux-mag.com/id/7564/#comment-7120 <p>Very informative!</p> <p>One question: when I cat /sys/block/sda/queue/scheduler it shows additional entries:</p> <p>$ cat /sys/block/sda/queue/scheduler<br /> noop fifo anticipatory [deadline] cfq vr </p> Very informative!

One question: when I cat /sys/block/sda/queue/scheduler it shows additional entries:

$ cat /sys/block/sda/queue/scheduler
noop fifo anticipatory [deadline] cfq vr

]]> By: laytonjb http://www.linux-mag.com/id/7564/#comment-7121 laytonjb Wed, 30 Nov -0001 00:00:00 +0000 http://www.linux-mag.com/id/7564/#comment-7121 <p>@speed145a</p> <p>Glad you liked the article. I\'m hoping to expand a little more in future articles to talk about how to tune various schedulers and how to measure the impact on workloads (probably just some benchmarks).</p> <p>I\'m not sure where the other two schedulers come from. A quick google didn\'t turn up too much. Can you tell us about your distro and perhaps what the system is doing?</p> <p>Thanks!</p> <p>Jeff </p> @speed145a

Glad you liked the article. I\’m hoping to expand a little more in future articles to talk about how to tune various schedulers and how to measure the impact on workloads (probably just some benchmarks).

I\’m not sure where the other two schedulers come from. A quick google didn\’t turn up too much. Can you tell us about your distro and perhaps what the system is doing?

Thanks!

Jeff

]]> By: speed145a http://www.linux-mag.com/id/7564/#comment-7122 speed145a Wed, 30 Nov -0001 00:00:00 +0000 http://www.linux-mag.com/id/7564/#comment-7122 <p>I also tried a google search without any luck :-)</p> <p>I\'m running Arch with the 2.6.31-zen1 kernel. I\'m sure it\'s something in the ZEN kernel... or perhaps something to do with the BFS in the kernel. </p> I also tried a google search without any luck :-)

I\’m running Arch with the 2.6.31-zen1 kernel. I\’m sure it\’s something in the ZEN kernel… or perhaps something to do with the BFS in the kernel.

]]> By: laytonjb http://www.linux-mag.com/id/7564/#comment-7123 laytonjb Wed, 30 Nov -0001 00:00:00 +0000 http://www.linux-mag.com/id/7564/#comment-7123 <p>@speed145a,</p> <p>I did some things around Xen but I wasn\'t sure. In particular i did find references to vr-sched.c but I didn\'t look at the code or find details of the scheduler. But I do think it\'s around Xen somehow but I don\'t know the purpose of the additional schedulers. Might be worth posting to an archlinux group.</p> <p>Jeff </p> @speed145a,

I did some things around Xen but I wasn\’t sure. In particular i did find references to vr-sched.c but I didn\’t look at the code or find details of the scheduler. But I do think it\’s around Xen somehow but I don\’t know the purpose of the additional schedulers. Might be worth posting to an archlinux group.

Jeff

]]> By: suresh17 http://www.linux-mag.com/id/7564/#comment-7124 suresh17 Wed, 30 Nov -0001 00:00:00 +0000 http://www.linux-mag.com/id/7564/#comment-7124 <p>Do we really need to run update-grub after modifying menu.lst (it\'s not lilo) Isn\'t? </p> Do we really need to run update-grub after modifying menu.lst (it\’s not lilo) Isn\’t?

]]> By: suresh17 http://www.linux-mag.com/id/7564/#comment-7125 suresh17 Wed, 30 Nov -0001 00:00:00 +0000 http://www.linux-mag.com/id/7564/#comment-7125 <p>BTW, very nice articles by Jeff on Linux Magazine.. Actually I have become a fan of you! Thanks Jeff! </p> BTW, very nice articles by Jeff on Linux Magazine.. Actually I have become a fan of you! Thanks Jeff!

]]> By: richarson http://www.linux-mag.com/id/7564/#comment-7126 richarson Wed, 30 Nov -0001 00:00:00 +0000 http://www.linux-mag.com/id/7564/#comment-7126 <p>A little too late, but it might be useful anyway :)</p> <p>It\'s not Xen but Zen: a kernel specifically patched to behave better on desktops, it includes several different patches and that\'s where the extra schedulers come from.</p> <p>Here are some URLs:</p> <p>http://www.zen-kernel.org/<br /> http://www.zen-kernel.org/about<br /> http://www.zen-kernel.org/included-code </p> A little too late, but it might be useful anyway :)

It\’s not Xen but Zen: a kernel specifically patched to behave better on desktops, it includes several different patches and that\’s where the extra schedulers come from.

Here are some URLs:

http://www.zen-kernel.org/
http://www.zen-kernel.org/about
http://www.zen-kernel.org/included-code

]]> By: richarson http://www.linux-mag.com/id/7564/#comment-7127 richarson Wed, 30 Nov -0001 00:00:00 +0000 http://www.linux-mag.com/id/7564/#comment-7127 <p>@suresh: it\'s a Debian (and Ubuntu, etc.) thing, you don\'t modify the boot parameters directly but some sort of a token (kopt) and then update-grub updates all your kernel entries. </p> @suresh: it\’s a Debian (and Ubuntu, etc.) thing, you don\’t modify the boot parameters directly but some sort of a token (kopt) and then update-grub updates all your kernel entries.

]]> By: suresh17 http://www.linux-mag.com/id/7564/#comment-7128 suresh17 Wed, 30 Nov -0001 00:00:00 +0000 http://www.linux-mag.com/id/7564/#comment-7128 <p>@richarson: ah, ok. Thanks! I wasn\'t aware of this. </p> @richarson: ah, ok. Thanks! I wasn\’t aware of this.

]]> By: grabur http://www.linux-mag.com/id/7564/#comment-7129 grabur Wed, 30 Nov -0001 00:00:00 +0000 http://www.linux-mag.com/id/7564/#comment-7129 <p>Great introductory article.</p> <p>Deadline here on Ubuntu 8.04.I\'d be interested to know the advantages/benefits of using specific schedulers, and how the Linux ones compare with other OS kernel schedulers, is Linux an innovator here? </p> Great introductory article.

Deadline here on Ubuntu 8.04.I\’d be interested to know the advantages/benefits of using specific schedulers, and how the Linux ones compare with other OS kernel schedulers, is Linux an innovator here?

]]> By: idallen http://www.linux-mag.com/id/7564/#comment-7130 idallen Wed, 30 Nov -0001 00:00:00 +0000 http://www.linux-mag.com/id/7564/#comment-7130 <p>Re: Using NOOP on an SSD.<br /> I would worry that the lack of write ordering in NOOP would mean *more* writes to the SSD, since only adjacent requests would be merged. The other schedulers would re-order all the outstanding writes and merge all the adjacent requests. </p> Re: Using NOOP on an SSD.
I would worry that the lack of write ordering in NOOP would mean *more* writes to the SSD, since only adjacent requests would be merged. The other schedulers would re-order all the outstanding writes and merge all the adjacent requests.

]]>