Few things to point out for people not actively paying attention every day:

https://issues.apache.org/jira/browse/CASSANDRA-13486 : Amitkumar Ghatwal
 and Rei Odaira from IBM have contributed some changes to make Cassandra
work properly on PPC hardware, including a new Row Cache implementation
that uses their proprietary flash hardware (
https://www-304.ibm.com/webapp/set2/sas/f/capi/home.html ) via a third
party, Apache licensed library. It seems like it'd be great if we could
bring it in tree. We have other vendor-specific (AWS snitch) and
vendor-contributed (GCP Snitch) libraries in tree, so this seems like a
reasonable goal (to me personally, as I editorialize). If you have strong
opinions about the right way to do this (or a good reason NOT to do this),
now may be a good time to share those.

https://issues.apache.org/jira/browse/CASSANDRA-13497 : Blake's added a
ticket for in-tree testing guidelines. These floated on the list a bit.
Nobody seemed to object. One step closer to formalizing testing methods.

https://issues.apache.org/jira/browse/CASSANDRA-12944 : The biggest ticket
nobody seems to notice (or at least, nobody seems to comment on) is
Stefan's Diagnostic Events ticket; this may go hand-in-hand with Blake's
testing ticket (#13497).

https://issues.apache.org/jira/browse/CASSANDRA-13474 : Pluggable storage
engine ticket. Has some high level features/goals, if you're deeply
invested in the storage engine/format, you may want to throw your 2 cents
in and make sure that whatever abstraction is selected supports the
features/requirements you care about.

Bookkeeping wise: all of the 3.11.0 "features and enhancement" tickets have
been bumped to 4.x, as the vote months ago was to make 3.11.0 bug-fix only.
The dtests have single-digit failures (2-5 each). If you're anxious to get
3.11.0 officially out the door, picking one of those 2-5 failing dtests and
fixing the test (or fixing the bug in the code) is the way to get it cut.

- Jeff

Reply via email to