[
https://issues.apache.org/jira/browse/CASSANDRA-4734?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13633224#comment-13633224
]
Edward Capriolo commented on CASSANDRA-4734:
--------------------------------------------
Anecdotal, I feel few actually understand CL, (heck maybe I do not even
understand it). What I find, is that users come to realize that Cassandra is
not atomic in the "atomic isolated" sense. Once that mental hurdle is crossed
users fall into a thinking similar to CL.ONE is the most likely to succeed, and
the fastest. Naturally since ONE is fastest and most likely to not cause
problems it becomes the application default.
IMHO there is a logic to setting CL at the column family level, however that
implies that all operations on a column family would require the same CL which
is not true, even inside a row key some columns might require a different CL
based on application requirements. Personally, I do not see how anything but a
per operation CL could be correct. If only 1/3 natural endpoints is up I would
not want my client to deny me the ability to write because the column family,
keyspace,...whatever dictates the CL it should be written at.
> Move CQL3 consistency to protocol
> ---------------------------------
>
> Key: CASSANDRA-4734
> URL: https://issues.apache.org/jira/browse/CASSANDRA-4734
> Project: Cassandra
> Issue Type: Task
> Components: API
> Reporter: Jonathan Ellis
> Assignee: Sylvain Lebresne
> Fix For: 1.2.0 beta 2
>
> Attachments: 0001-Move-consistency-level-to-the-protocol-level-2.txt,
> 0001-Move-consistency-level-to-the-protocol-level-3.txt,
> 0001-Move-consistency-level-to-the-protocol-level.txt,
> 0002-Remove-remains-of-4448-3.txt, 0002-Remove-remains-of-4448.txt,
> 0002-Thrift-generated-file-diffs-2.txt,
> 0003-Thrift-generated-file-diffs-3.txt, 0003-Thrift-generated-file-diffs.txt
>
>
> Currently, in CQL3, you set the consistency level of an operation in
> the language, eg 'SELECT * FROM foo USING CONSISTENCY QUORUM'. It now
> looks like this was a mistake, and that consistency should be set at
> the protocol level, i.e. as a separate parameter along with the query.
> The reasoning is that the CL applies to the guarantee provided by the
> operation being successful, not to the query itself. Specifically,
> having the CL being part of the language means that CL is opaque to
> low level client libraries without themselves parsing the CQL, which
> we want to avoid. Thus,
> - Those libraries can't implement automatic retries policy, where a query
> would be retried with a smaller CL. (I'm aware that this is often a Bad
> Idea, but it does have legitimate uses and not having that available is seen
> as a regression from the Thrift api.)
> - We had to introduce CASSANDRA-4448 to allow the client to configure some
> form of default CL since the library can't handle that anymore, which is
> hackish.
> - Executing prepared statements with different CL requires preparing multiple
> statements.
> - CL only makes sense for BATCH operations as a whole, not the sub-statements
> within the batch. Currently CQL3 "fixes" that by validating the given CLs
> match, but it would be much more clear if the CL was on the protocol side.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
