Spurious high disk io rate alerts from remote server

Question

I have a server running on linode. The server is a headless Ubuntu 10.0.4 64 bit installation.

I have recently started getting high disk I/O alerts from the server. As far as I know, nothing much has changed on the server.

Is there a *nix utility that I can run on the server to either monitoring and log, or perhaps analyze an existing log in order to find out which process(es) are "trashing" the disk?

Answer 1

some ideas:

IOtop: http://guichaz.free.fr/iotop/

Atop: http://www.atoptool.nl/

Nmon: http://www.ibm.com/developerworks/aix/library/au-analyze_aix/

Collectl: http://collectl.sourceforge.net/Process.html