Details
-
Bug
-
Status: Resolved
-
Normal
-
Resolution: Fixed
-
None
-
None
-
Normal
Description
In a moment of weakness and for somewhat weird reasons I broke the
serialization of CFMetaData ids. It's even a double failure on my part since I
should have looked while reviewing CASSANDRA-9709 why we had this weird
reversion of bytes in the first place and figure out it was my mistake, but I
have been lazy instead.
Anyway, the bytes of the table id are currently reversed in trunk which
shouldn't be. This is actually one of the reason for the failure of
CommitLogUpgradeTest (so we do have a test catching this), but there is
other reasons for these failures (more related to the test and which I'll
handled on CASSANDRA-9790) and I'd like to get this one fixed asap, hence the
separate ticket.