Bret McGuire created CASSJAVA-99:
------------------------------------
Summary: See if we can improve on DefaultCodecRegistry.CacheKey
Key: CASSJAVA-99
URL: https://issues.apache.org/jira/browse/CASSJAVA-99
Project: Apache Cassandra Java driver
Issue Type: Improvement
Components: Core
Reporter: Bret McGuire
Came up in conversation around [the recent
PR|https://github.com/apache/cassandra-java-driver/pull/1988] to minimize
excess object allocation from varargs usage in the same code. As part of that
conversation it was pointed out that we're creating CacheKeys for every
addition to the cache {_}and for every lookup{_}. We need to do a lookup for
every field of every row we're decoding so there's _a lot_ of object churn
coming out of this.
The complication here is that we need to look up codecs based on three
independent criteria:
* The underlying CQL type
* The target Java type
* The question of covariance
This is currently implemented via CacheKey with consideration of all three of
these values in hashCode(). But while working on this ticket I couldn't help
but think that there has to be a better way to manage this lookup. This ticket
is intended to address that issue.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
