When TVDB.com is offline like it was tonight, does not fail very gracefully.
Instead it starts throwing NPE's all over the logs for each attempt to get the
season. Because it continues to fail to connect, each call to the TVDB API
blocks until the HTTP timeout (around 3 seconds I think). This is painfully
slow to experience.
See the logs here:
http://pastebin.com/Whsn3WeD
Recommendation:
Check for TVDB.com being online. If it does not provide an adequate response
during the initial setup then take an action there. The action should be
configurable. Maybe to fail to an alternative service or to allow it to run
without the additional info from TVDB.
When the TVDB service goes down in the middle of a script executing the above
solution would not work. In that scenario we should detect TVDB failure
patterns and take the same configurable action mid processing when a TVDB
failure pattern is detected. An example pattern could be more than x failed
responses within y minutes, where x and y are configurable as well.
Original issue reported on code.google.com by robbiene...@gmail.com on 24 Aug 2012 at 2:09
Original issue reported on code.google.com by
robbiene...@gmail.com
on 24 Aug 2012 at 2:09