Past

Event scavenger was created out of a need to monitor multiple servers in a production environment and have a view what is happening across all the servers at the same moment of time. It began somewhere in 2006 and was continuously updated with new functionality as needed. It also provided a way to pull reports and stats from a database which is not so easy from event logs directly - this would have involved exporting the logs to a database of spreadsheet anyway.
Version 1 and 2 didn't have much more functionality than the current version 3 - merely optimizations and incremental improvements. The only reason version 3 was created was to move the source to CodePlex plus having an excuse to clean up the code and get rid of old 'stuff' that wasn't used anymore.

Future

There are a couple of things that can be improved or even stripped from the system depending on need. For example, at the moment the archiving functionality is really simple and not too useful - EventLogHistory table. This might be stripped or changed at some time.
One thing that is still being considered is changing the algorithm of the collector (service) to rather spawn threads as needed and not having a separate thread per event log polled like present. Most of the time threads are sitting idling anyway. However, the overhead of idling threads seems minimal at the time and they don't take much resources. The current algorithm has been used for about 4 years and has never given any performance or resource issues on any of the machines it was used on. The only time performance could be an issue is when adding a whole lot of new event logs that must be imported from scratch. This takes some time to complete.

Last edited Apr 28, 2010 at 5:39 AM by RudolfHenning, version 1

Comments

No comments yet.