Comments on: Tuning CFQ – What Station is That? http://www.linux-mag.com/id/7572/ Open Source, Open Standards Sat, 05 Oct 2013 13:48:18 +0000 hourly 1 http://wordpress.org/?v=3.1 By: sgtaaron http://www.linux-mag.com/id/7572/#comment-7152 sgtaaron Wed, 30 Nov -0001 00:00:00 +0000 http://www.linux-mag.com/id/7572/#comment-7152 <p>The grammar in this story is terrible. I read the previous article just now and note that the grammar in that story is bad, too. Does anybody at Linux Magazine bother to proofread their stories? </p> The grammar in this story is terrible. I read the previous article just now and note that the grammar in that story is bad, too. Does anybody at Linux Magazine bother to proofread their stories?

]]> By: porridge http://www.linux-mag.com/id/7572/#comment-7153 porridge Wed, 30 Nov -0001 00:00:00 +0000 http://www.linux-mag.com/id/7572/#comment-7153 <p>Wow, what a useless article.. should have finished after listing the knobs, if tweaking them didn\'t bring any improvements. </p> Wow, what a useless article.. should have finished after listing the knobs, if tweaking them didn\’t bring any improvements.

]]> By: laytonjb http://www.linux-mag.com/id/7572/#comment-7154 laytonjb Wed, 30 Nov -0001 00:00:00 +0000 http://www.linux-mag.com/id/7572/#comment-7154 <p>@porridge,</p> <p>Sorry - but a \"no\" answer can be as good as a \"yes\" answer. At least I showed for the test configuration I used and the tests I ran that I saw little improvement. Guess what - at least I now know. I doubt you could have answered that question before the testing.</p> <p>Plus, I hope you understand that you can make tweaks but you don\'t know if they help or not, until you actually test them. Plus you should test them with applications that have some meaning to you and you situation. </p> <p>If I had found a \"magic\" tweak combination would that have meant anything to you? Maybe or maybe not. Is your hardware the same? Did you run the same tests? Are the tests meaningful to you? Sometimes coming up with a magic tweak doesn\'t help all situations.</p> <p>I\'m hoping that people reading the article will learn from it. Obviously you didn\'t. Remember the Thomas Edison phrase, \"If I find 10,000 ways something won\'t work, I haven\'t failed. I am not discouraged, because every wrong attempt discarded is another step forward\"</p> <p>Jeff </p> @porridge,

Sorry – but a \”no\” answer can be as good as a \”yes\” answer. At least I showed for the test configuration I used and the tests I ran that I saw little improvement. Guess what – at least I now know. I doubt you could have answered that question before the testing.

Plus, I hope you understand that you can make tweaks but you don\’t know if they help or not, until you actually test them. Plus you should test them with applications that have some meaning to you and you situation.

If I had found a \”magic\” tweak combination would that have meant anything to you? Maybe or maybe not. Is your hardware the same? Did you run the same tests? Are the tests meaningful to you? Sometimes coming up with a magic tweak doesn\’t help all situations.

I\’m hoping that people reading the article will learn from it. Obviously you didn\’t. Remember the Thomas Edison phrase, \”If I find 10,000 ways something won\’t work, I haven\’t failed. I am not discouraged, because every wrong attempt discarded is another step forward\”

Jeff

]]> By: nouiz http://www.linux-mag.com/id/7572/#comment-7155 nouiz Wed, 30 Nov -0001 00:00:00 +0000 http://www.linux-mag.com/id/7572/#comment-7155 <p>Do someone know something similar to the nice program for the cpu, but for the disk?</p> <p>It happen from times to times that I must do some very heavy io operation on a nfs server(like du -ch --max-depth 1 ...). I do it locally, but this slow down all user of this nfs server. I don\'t see any way to do this in all the IOScheduler that you presented or in how we can tune them.</p> <p>Thanks</p> <p>Frédéric Bastien </p> Do someone know something similar to the nice program for the cpu, but for the disk?

It happen from times to times that I must do some very heavy io operation on a nfs server(like du -ch –max-depth 1 …). I do it locally, but this slow down all user of this nfs server. I don\’t see any way to do this in all the IOScheduler that you presented or in how we can tune them.

Thanks

Frédéric Bastien

]]>