Kyle Blaney wrote:

I'm looking for help because our slapd service occasionally terminates unexpectedly on Windows. I know that Windows isn't the ideal platform and we're moving everything to Linux eventually but we need to put up with Windows for a short while longer. Our configuration is OpenLDAP 2.3.39 with BDB 4.4.20 (although we've also seen the same problem in OpenLDAP 2.3.38 and 2.3.41) built using MinGW/MSYS.

The problem is not readily reproducible but it has been seen on many different systems. I don't have any further details because we've set the service's recovery options to automatically restart the service when it terminates unexpectedly. All reports indicate that the service is not doing anything in particular when it unexpectedly terminates. For exampe, it's not being heavily accessed. In fact, it often stops when it's not being accessed at all. It runs perfectly well for hours or days and then just stops before being automatically restarted. The only indication that it even stops is an error logged in the Windows Event Viewer by the Service Control Manager that says "The slapd service terminated unexpectedly. This has happened 1 time(s)".

Are there any specific logging options I should turn on so that I can provide more details as to what the service is doing just before it terminates?

Typical logging facilities won't tell anything. The way to determine what happened is to run slapd from a command shell, instead of as a service, with minimal debugging enabled (e.g. -d none). Ideally, run it under gdb and let gdb report on how it terminates.

Is this likely to have the same root cause as ITS#5469 (Assertion failure causes slapd crash): _http://www.openldap.org/its/index.cgi/Software%20Bugs?id=5469;selectid=5469_...

Possible. You should be able to download 2.3.42 and verify if that's the case.

Then again, remember the Windows haiku:

yesterday it worked today it is not working Windows is like that