| PLEX86 | ||
|
The cost of copying 1462^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Saw that and freaked! Am I majorly losing performance? disk check on every bootup On Saturday 27 May 2006 23:50, christine stood up and spoke the following words to the mbuttes incomp.os.linux.misc...: If you have... ~# hdparm -c 1dev-hdadev-hdc dev-hda: setting 32-bit IOsupport flag to 1 IOsupport = 1 (32-bit) dev-hdc: setting 32-bit IOsupport flag to 1 IOsupport = 1 (32-bit) 65536+0 records in 65536+0 records out user 0m0.270s sys 0m49.630s ~# hdparm -c 3dev-hd{a,c} dev-hda: setting 32-bit IOsupport flag to 3 IOsupport = 3 (32-bit w-sync) dev-hdc: setting 32-bit IOsupport flag to 3 IOsupport = 3 (32-bit w-sync) 65536+0 records in 65536+0 records out user 0m0.280s sys 0m51.120s Not much of a difference... (previous 16-bit mode was 27.8 MB-s) But on the cp -r test, I now get: ~$ top top - 13:47:55 up 1:14, 2 users, load average: 2.03, 0.89, 0.50 Tasks: 52 total, 3 running, 49 sleeping, 0 stopped, 0 zombie Cpu(s): 0.3% us, 18.6% sy, 0.0% ni, 0.0% id, 79.4% wa, 1.7% hi, 0.0% si I hit 'q' when load average fell back to 2.03, in the previous test load average was still climbing slowly when I hit 'q'. ~# mountdev-hdc10mnt-hd~# mkdirmnt-hd-temp ~# time $(cp -rhome-install-slackware-10.1-mnt-hd-temp-; sync) user 0m0.320s sys 0m56.410s ~# du -shmnt-hd-temp2.9G mnt-hd-temp~# findmnt-hd-temp-* wc -l 6203 So a 'hdparm -c 3dev-hd{a,c}' does make a small improvement in load average, worth trying? Grant. -- ... The computer scientist, who had listened to all of this said, "Yes, but where do you think the chaos came from?"
|
||||
Linux groups from Newsgroups The #1 Usenet Provider on the Internet
|
||||