- Author
- Download
- (direct download)
- SVN
- (svn url)
- Docs
- (docs)
- License
- GPL Public License
- for Tivoli Storage Manager
- Documentation: http://blackcat.ca/svn/trunk/tsmdailyrep/docs/index.html
See the sample report for an idea of that this program does.
TSM is a data management solution that is offered by IBM. It’s a huge, sprawling, flexible monstrosity that work perfectly… if you watch it and keep it fed and watered. TSM is also very scriptable and extensible.
This program attempts to give a TSM administrator a daily report that contains all the things that need to be periodically looked at. Wherever possible it will offer advice on how to fix problems found or attempt to resolve them itself. It will generate a nice email with the report and will modify the subject to flag and problems found so that you know when it’s really important to have a peek inside the report.
If you know of anything that you’d like added, you find a bug, or just can’t get it to run for you, please email me at paudley@blackcat.ca and I’ll see what I can do to help. Please Note: I no longer have regular access to a TSM system for testing, if you’d like troubleshooting I can probably work something remotely.
Thanks to the many people that have passed along code or suggestions and to the Wellcome Trust for funding my time at the University of Dundee.
Change History
v1.0
- prelimary version, ugly perl :)
v1.1
- refactored perl code
- summarize more log messages
- make scratch queries actually work on other systems
- added checks for:
- nodes that haven’t talked to the server in a while
- filespaces that haven’t been backed up in a while
- added reports for:
- average fill of tapes in each device class
v1.2
- add session section
- show loaded/empty status for drives
- catch restored objects in the client summary and warn for unknown types
- check for MIME::Lite and disable email if not found
- revamp configuration sections, use more hashes
- add command line arguments for many things
- add fine-grained section control
- added the alter_library flag to toggle commands that might alter TSM
- added the option to skip slow or iterative queries
v1.3
- add reporting of drives that need to be cleaned in the drives section
- significant activity log parsing speedups for very large logs
- create a serious errors section
- detect data-integrity errors and attach a list of them if found
- informating about audited volumes
- ability to audit random volumes
- added memoize support and the no_queries options to help in debugging and running offline.
v1.4
- make alerts appear in the serious errors section
- show breakdown of volumes by type per library in the volume information section
- make alter_tsm a command line option
- try to work around a missing Pod::Usage
- change the “use warnings” pragmas to just a “-w” at the top of the script
- display locked client nodes
- add column numbers to debug_tables output
- add db and log mirroring information
- add an advice function
- report private volumes that have no recorded last use, optionally turn these into scratch volumes.
- fix labeling of bulk io volumes
- add data transfer time to client display
- detect slow nodes
- add list of added users and nodes, also show policy domain association
- add config file parsing
v1.5
- add a schedule graph section
- exclude nodes that you don’t care about
- added the script information section
- added a debug section — mainly for me while developing
- report SQL errors
- support narrowing report to specific policy domains
- detect nodes that were terminated for being idle
