I just solved an problem with email archive. The service was stopping for no reason that I could detect, and didn't post any messages to the event log. I solved the problem by restarting the server, so I blame my server, not the Laserfiche software for the fact that it failed.

What did concern me is how it behaved before I restarted the server. Every time I started the service, it would delete one message from the monitored folder, and not do anything with it. The service was configured to move the message to another email folder after reading it, as well as loading it into my repository.

It seems that it should be possible to architect the service so that no failure would ever delete a message before putting it somewhere else. If a failing server causes the the service to misbehave I would far rather have duplicate information than missing information. Duplicate information is generally both detectable and recoverable. Missing information is neither.