Hi all, new version is out... Most of the changes is on the fork functions so please upgrade to this version because it solve some problems with the forking of process.
For netapp users the cpu report is now using the average cpu used instead of the total and the script will report as data to hobbit all the stats performance data. The included patch to the hobbit server will allow the user to save all those data to rrd file. Definition of those graphs are on the way, I'm currently working on them and on a better documentation.
For people using Perl 5.8.0 (for example red hat 3 users) they have to upgrade the Sys::SigAction module to version 0.07 because the 0.06 version have a bug that will make the script exit without reporting anything.
This is the complete changelog:
Hobbit_fd_lib.pm v1.14
- Added function for uptime checking
- Added an option to change the dbcheck/netapp/beastat reporting column under the hobbit server to report warning or error o nly if there are more than a defined number of errors
- Completely changed the way on how child fork and how they'll pass data to the parent process
- The program will now exit if the perl version is 5.8.0 and Sys::SigAlert version is less then 0.07 because the module will not work in that configuration
- Corrected a bug in a time calculation function not working with some dates and creating a lot of warning
Beastat.pl v1.06
- Added an option to have the uptime in a different column (see the beastat.ini file for options)
- Corrected a bug with the uptime reporting
DbCheck.pl v1.07
- Added an option to have the uptime in a different column (see the dbcheck.ini file for options)
- Corrected a bug to get the default value for tblspace check
- Corrected a bug in the check for errors in the valtocheck parameter
- Corrected some bugs on the locks report queries
Netapp.pl v1.10
- Added an option to have the uptime in a different column (see the netapp.ini file for options)
- Added an option to check and report for timedifference between the hobbit server and filers
- Added the report of extended performance data gathered via stats command (those data will be collected into rrd file after patching the hobbit source and will report about many statistical counters. To know what can be graphed check the stats command and in particular stats list counters with object: aggregate, volume, disk, qtree, iscsi, fcp, cifs, lun, nfsv3, ifnet, processor, system
- Corrected cpu statistics gathering via sysstat -m on multiprocessor platform using the average cpu time instead of the total (better reporting of cpu performance).
- Full report on cpu information with total, average and single processor use on the cpu column
- Some bug fixed to various functions