With jobs as processes, it is by design that WConfD can be restarted while jobs are running. So jobs need to be able to handle a connection to WConfD not working and be able to retry some times. On top of that, if things really go wrong, we must not end up in the state where the queue dies but the top-level thread (and hence the process) survives. This will hopefully solve itself when jobs become single threaded, but for the time being we have to be careful with assertions not valid any more.
Klaus Aehlig (2): Retry connections to WConfD Remove obsolete assertion lib/jqueue/__init__.py | 13 ++++--------- lib/wconfd.py | 18 +++++++++++++++++- 2 files changed, 21 insertions(+), 10 deletions(-) -- 2.0.0.526.g5318336
