It's a database error alright, but only in the sense that the database process appears to have gone away somehow, since the substance of the error is the "Connection has terminated" bit. If that's the case, the process reporting the error isn't the one that really had the problem, so the error message doesn't convey a lot about the underlying cause. Since there haven't been other reports of this I can recall seeing, it's not something I've dug into before and I can't suggest a resolution until I can investigate what the real root cause might be.
One thing that might help is to confirm whether, at the point that the above message appears, the rteng9.exe process (which is the SQLAnywhere run-time used by the console and Configuration Server) is still running or not. If it is no longer running, then it's a matter of me figuring out why. It would also help to confirm which version of GSS you are experiencing this in, as well (whether GSS2.5 or one of the older editions).
internally, the WMI data is processed one machine at a time, with that inventory processing split between two threads in internal queues, so the number of machines involved at once shouldn't be a direct contributor to the cause, but should be a more indirect factor (e.g. if the root cause is some low-probability event, or some kind of temporary load spike). One thing that does strike me about this, though, is that the specific point in the code where that INSERT is taken place is immediately preceded (in the same transaction) by a query to remove the existing data in the database for a particular machine, which gives me pause to wonder if there's some kind of resource consumption problem triggered by that (since that delete query will be doing a LOT of work).
I'll have a think about how to investigate this, since this isn't something we've seen happen here and the underlying problem could be in in third-party code making it tricker to diagnose than most problems. My initial feeling is that I'd probably want to start by getting hold of your console machine database just to inspect and see how the database engine responds when processing your data to see what kind of load it's under when doing this. It would be a good idea if you could get a backup of your database
using the backupdb.wsf script and get in touch with me by e-mail (use the address in my Connect forum profile) so we can arrange that.