Performance impact of fschange on C:\windows\* ?

[Jason]

Has anyone put into production an input stanza that runs an fschange on all of C:\windows?

A) what is the performance impact, if any, that was noticed on the host running this fschange?

B) This will be run on a light forwarder. If anyone is doing C:\windows fschange on a light forwarder successfully (despite the SPL-25294 bug) please let me know!

[gkanapathy]

If you just want changes, it may be much better to enable Windows FS auditing and use Splunk to monitor the Windows Security Event Log for changes instead. The overhead is much lower, and you will catch changes that Splunk will not catch, as the audit is generated by the filesystem itself rather than by polling.

[Jason]

As for testing, nope. The client wants it, but I don't know if we have access to any "typical" servers that are not production. I doubt my laptop would be a similar enough environment to their servers.

[Jason]

True, I assume they don't want to index the files, so let's go with that.

[Lowell]

Have you tried it on any of your machines? Seems like something like this would be worth the effort of setting up a standalone test. That's my 2 cents. BTW, I'm assuming that you are not going to be indexing any of the files, you are just looking for changes, right? (in other words fullEvent=false), you question doesn't indicate either way.

[Jason]

For reference, C:\windows just on my laptop is 5.2 GB, almost 25,000 files, and 2100+ folders. Who knows how often files in it are changed.