Stability improvements

A lot of work has been done the last few weeks, but you might not have noticed. Most time has been spent reducing complexity, which is always a good thing. Quite some code was shared between the GUI, the public website and the API. Also, test code was running from a different directory instead of from a different url, making it necessary to have some nasty config checks in the main code.

That has all changed now. The source code has been refactored and now all parts have proper separation. On top off that ShadowTrackr now uses a Continuous Integration pipeline: on every code check-in we run automated sanity checks and tests.

New domain alerts

The migration from traps to alerts is now completed, and we’ve added some new alerts for domains. You can now choose to set alerts for:

• Whois domain available
• Whois domain registered
• Whois record found
• Whois domain changed registrar
• Whois record changed
• Whois record expired
• Whois record expires next week
• Whois record expires in two weeks
• Whois record expires in three weeks

Development time in the coming period will be spent on adding more configurable alerts, expanding the use for tags, and hunting for bugs. If you see any bugs or other annoying things, please report them!

Naughty node - database error

Yesterday some client reported problems using the Shadowtrackr webinterface. After some initial patchwork the root cause was found: a faulty database node.

The transaction logs one this specific node had grown too big, and were somehow not rotated. This resulted in a full disk, and errors for all queries sent to this database node. If you ended up on this node yesterday, you might have experienced problems logging in or viewing data.

Overnight I’ve rebuild the database node an have taken measures to make sure this doesn’t happen again. I apologize for all the trouble.