Re: High iowait
Lothar Braun <[email protected]> Tue, 22 Aug 2006 11:38:10 +0200
| Newsgroups | org.kernel.vger.linux-config,org.kernel.vger.linux-admin |
|---|---|
| Message-ID | <[email protected]> |
--nextPart6058115.Ur7apUI16p Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline Hi, On Tuesday 22 August 2006 04:31, [email protected] wrote: > I have a storage raid attached to my linux server running CentOS 4.3 > kernel 2.6.9-39.0.2.ELsmp. My server goes into huge iowait when i try > to read/write lot of data to the storage device. How can i debug > whether the problem is with the storage device or some setting of the > operating system. Is there any known issue with this kernel? How can i > tune my system to reduce iowait time? Why should you want to reduce the iowait time? Iowait, as shown by top, mea= ns=20 afaik that there are some processes waiting for io to complete and nothing= =20 else to do for the processor. If you're running some CPU eating processes while your I/O operations are=20 performed, then the iowait value should go down (this is because your CPU c= an=20 then do more sophisticated work than waiting) Regards, Lothar --nextPart6058115.Ur7apUI16p Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.3 (GNU/Linux) iD8DBQBE6tCHCllgm08lUqkRAp3mAJ9HzsA/zLfTSwSM7bwOP6k9UmHw9QCfREzB vwobV+P8Jhv8IaVh8DnRJ+E= =oi2c -----END PGP SIGNATURE----- --nextPart6058115.Ur7apUI16p--