Well, unfortunately this could be a myriad of issues. First, have you gotten a NSD? There are tools to help read them, but opening a ticket with IBM and getting them the NSD to help troubleshoot is the first and quickest way to get to the bottom of this. Second, I’d need to know more about your environment, but if you have multiple servers, would it be possible to put a fresh replica of admin4.nsf on the server that is crashing? You need to figure out what adminP is trying to process when it dies. The ease of that depends on how much activity is in your admin4.nsf. If there are not many things being processed in AdminP, you might be able to pinpoint it easily. If you have a large environment, this could be more problematic.
Please post some more information about your environment, and we’ll see what we can do to help.