Osm-logical: sanity checks about osmdbt HOT 2 CLOSED

I don't think the added complexity is worth it here. Either these are errors which can actually happen in real life in which case we must handle them. Or they are not, which I think is the case here. But then doing the simplest thing is enough to get at least something when testing or so. And that's what we have. We can spend forever coming up with scenarios how this can all fail, but at some point we have to ask ourselves whether it is worth it.

from osmdbt.

I see that those errors are currently printed to stderr, so they're no totally lost:

        } else if (parts[2] == "X") {
            std::cerr << "Error found in logfile: " << line << '\n';

I think the basic question here is, what would a sysadmin do with those messages? Would they be part of some monitoring, or disappear into void?

At least the elog warnings would also appear as Warning in the Postgresql log files, where they might be easier to consume for an automated monitoring. Unfortunately, I'm not aware of any guidelines here, so this may all a bit theoretical after all.

from osmdbt.