cassandra-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jeff Jirsa <>
Subject Almost sorta weekly cassandra wrapup
Date Sat, 22 Apr 2017 00:41:07 GMT
Few things worth mentioning in the cassandra world this week:

1) If you somehow missed the cassandra-on-the-rocks(db) thread, you should
check it out:
. FB/Instagram are proposing pluggable storage engines to make using
alternate backends (like RocksDB) possible/supported. This is common in
other databases (think MySQL with MyISAM+InnoDB), but is new to Cassandra.
If you have a strong opinion, you should reply to that thread.

2) There's a ticket on trying to be more clever with storage/replication
factors @ , that's
probably interesting if you're storage-bound.

3) The diagnostic events ticket ( ) has some
interesting tests added (
). Maybe that's a month old, but it's an interesting concept.

Corentin asked for a reviewer in IRC for:

And since he asked, I figured other people may want some reviews too.

Here are a few others that are PA and no reviewer: (backporting a
logging change from 3.6) (snapshotted
manifests are wrong) (rpm fixes) (rpm fixes) (using aws EFS will
fail on startup due to long overflow)

Less LHF'y, but potentially very important: (incorrect results
when including the same column in cql twice) (nodetool refresh
maintains levels) (CAS inconsistencies
on read - also includes a lot of CAS testing) (commitlog replay bug) (streaming for bulk

- Jeff

  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message