Am Mittwoch, 29. August 2012 schrieb Stan Hoeppner: > On 8/28/2012 2:01 PM, Martin Steigerwald wrote: > > Hi Stan, > > > > Am Montag, 27. August 2012 schrieb Stan Hoeppner: > >> For a desktop user workload, there will be no noticeable performance > >> difference, because such applications don't do parallel IO. > > > > Are you sure about 1) desktop applications I/O behavior and 2) NCQ? > > > > 1) I see noticeable difference for my Intel SSD 320 with different > > iodepth values in fio job. > > Shall we now split hairs, and have a 500 post discussion WRT the > definition of "desktop user workload"? > > If you consider fio, iozone, bonnie++ etc to be part of a typical > "desktop user workload" then you're simply out of touch with the real > world Martin. ;) > > Do you have a father/mother brother/sister wife/children? Do any of > them run fio on a regular basis? These people represent the typical > "desktop user workload". Sysadmins, devs, power users, do not.
While a typical desktop workload may not induce anything near an IO depth of 64, I still think several desktop applications accessing the disk at once can have the Linux kernel block layer use an iodepth > 1 on the disk. So maybe testing with I/O depth upto 10 or so make sense even for desktop workloads. -- Martin 'Helios' Steigerwald - http://www.Lichtvoll.de GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7 -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/201208291607.06354.mar...@lichtvoll.de
