CASSANDRA-19802: Harry Simulation test halts the JVM when errors are detected which loose all history in CI #3441
Conversation
There was a problem hiding this comment.
this is for consistency... having this test do seeds differently than the other tests makes it harder for people to jump in and see what's going on.
There was a problem hiding this comment.
the error is consistent for other tests, so migrating to that handling... right now the seed is lost most of the time, so this makes sure it is captured 100% of the time
There was a problem hiding this comment.
this is lost in CI....
build.xml
Outdated
ifesdjeen
left a comment
ifesdjeen
left a comment
There was a problem hiding this comment.
One nit, feel free to ignore. +1 LGTM!
There was a problem hiding this comment.
Should we turn into a suppressed error as above here, too?
There was a problem hiding this comment.
suppressed error
why would we want to suppress? Something unexpected happened and we have no clue how to handle... so we should fail the test. This might be the AssertError we now generate, but we have to push that up... but what if its a different error?
…ose all history in CI patch by David Capwell; reviewed by Alex Petrov for CASSANDRA-19802
dcapwell
force-pushed
the
CASSANDRA-19802
branch
from
f907b6a to
8d076d0
Compare
2 participants
No description provided.