launchpad-dev team mailing list archive

[Maris Fogels <maris.fogels@xxxxxxxxxxxxx>]

Hi all,

Question for the room: should a test case leave behind OOPS reports in 
/var/tmp/lperr.test?  Would detecting that a test left behind OOPS reports when 
it was not supposed to increase the quality of our code?

I discovered this while trying to debug the ec2 test suite hangups.  At the 
point where the test suite freezes there are 2200+ OOPS reports in lperr.test/. 
 Here is a sample:


75284.T309:Exception-Value: [Errno 12] Cannot allocate memory
76110.T865:Exception-Value: default __new__ takes no parameters
76123.T872:Exception-Value: http://evil.people.com/
76124.T873:Exception-Value: not a url
76125.T877:Exception-Value: None isn't acceptable as a value for Product.owner
78601.T1155:Exception-Value: Bug #504291: Store left in a disconnected state.
78601.T1156:Exception-Value: server closed the connection unexpectedly This 
probably means the server terminated abnormally before or while processing the 
request.


Maris

P.S. Pet theory: leaking OOPSes may help us pinpoint bugs that arise in the test 
environment itself.  Then perhaps my search for the source of the test suite 
hangs would not feel so much like looking for a needle in a haystack.