SURVIVOR: disk Check Module
About disk
basics
Module Type Scheduler Remote
check No Yes
syntax
     module disk {
       [warn <number>]
       [prob <number>]
       [filesystem <directory>]
       [filesystem ...]
       [diskcf <file>]
     }
     via <transport>
 

Dependencies

  • None

Arguments

Name Type Value
warn optional number between 0 and 101, default is 101 Percentage of disk in use to generate a warning.
prob optional number between 0 and 101, default is 101 Percentage of disk in use to generate a problem.
filesystem optional directory list Filesystems to examine.
diskcf optional file A configuration file containing filesystems and threshold to use.

Description

The disk module determines how close to capacity local filesystems are on the host to be checked. It uses df to determine the amount of capacity in use for each local filesystem.

warn and prob set, respectively, the percent of capacity that must be reached for a warning or problem to be generated. 101 will never match, and so can be used to suppress alerts at that level. If no filesystem is provided, all local filesystems are examined, otherwise the requested filesystem(s) only are examined.

If diskcf is specified, then warning and problem thresholds are read from the specified file, which must be installed locally on the host to be monitored, not on the scheduler host. The contents of the file are lines of the form

       fsregex,warn,prob
 
fsregex is a regular expression describing the filesystem names to match. These are real regular expressions and not globs, so to match all filesystems either ^/.* or .* must be used. To match all filesystems beginning with the letter 'u', use ^/u.*. To match all filesystems containing the string 'data', use data. Lines beginning with a # symbol are treated as comments. If diskcf is specified but a local filesystem does not match any lines in it, the argument values warn and prob will apply.

It may be useful to set result text significant for disk checks. This will clear an existing acknowledgement when the summary text changes, for example when a filesystem's usage changes, or a different filesystem begins to fill up.

Examples

  1. To generate a warning when any local filesystem reaches 80% of capacity and a problem at 90%:
         module disk {
           warn 80
           prob 90
         }
         via remote
         
  2. As above, but to only examine the filesystem /var:
         module disk {
            warn 80
    	prob 90
    	filesystem /var
         }
         via remote
         
  3. To use a local configuration file, defaulting to warnings at 90% and no problems for filesystems not found within:
         module disk {
           warn 90
           diskcf /etc/survivor/disk.cf
         }
         via remote
         
  4. To clear acknowledgements when the summary text changes:
         module disk {
           warn 80
           prob 90
         }
         via remote
         result text significant
         

$Date: 2006/11/19 18:51:10 $
$Revision: 0.6 $
keywords
diskcf
filesystem
prob
warn