cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Aleksey Yeschenko (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (CASSANDRA-9712) Refactor CFMetaData
Date Mon, 03 Aug 2015 18:33:05 GMT

    [ https://issues.apache.org/jira/browse/CASSANDRA-9712?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14652224#comment-14652224
] 

Aleksey Yeschenko commented on CASSANDRA-9712:
----------------------------------------------

bq. CachingOptions had two allowed syntaxes: map and (deprecated) string. Is it intentional
that the string syntax is removed? These changes should be included in NEWS.txt.

It's been deprecated since 2.1, just not documented. I added a NEWS.txt entry about the removal,
and opened CASSANDRA-9964 to fix the documentation oversight.

bq.  Deprecation of CompressionParams’ SSTABLE_COMPRESSION and CHUNK_LENGTH_KB, should be
mentioned in NEWS.txt, too.

Had been done a while ago in the original CASSANDRA-8384 ticket.

bq. CompactionParams.validate(): catch of InvocationTargetException and IllegalAccessException
should include the original cause to ConfigurationException.
(CompressionParams.validate() does that.)

Made them behave exactly in the same way.

bq. ColumnFamilyStore.setCompactionStrategyClass() seems not to actually set the new compaction
strategy in CFS.
ColumnFamiliyStore.setCompressionParameters() should validate the params before these are
actually applied.

Made the method throw {{UnsupportedOperationException}} for now, since it's broken even without
CASSANDRA-9712. CASSANDRA-9965 will fix it properly.

bq. LongCompactionsTest.defineSchema(): compactionOptions.put() is redundant (that test fails
for this reason)

Fixed, thanks.

bq. Utests look good (except the one mentioned above). Dtests run failed.

{{caching}} syntax was the issue. [~philipthompson] fixed the dtests, thankfully.

bq. SpeculativeRetryParam: remove none() + all() (or just return the constant) - but I'd be
fine with leaving it as it is.

Left them be for consistency with other *Params

bq. CompressionParams: use „public static final“ instead of „public final static“
(and for private/volatile)

Fixed.

bq. ReplicationParams.CLASS is not final

Fixed.

Thanks for fast feedback. I'll rerun cassci tests now and see where it gets us.

> Refactor CFMetaData
> -------------------
>
>                 Key: CASSANDRA-9712
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-9712
>             Project: Cassandra
>          Issue Type: Improvement
>            Reporter: Aleksey Yeschenko
>            Assignee: Aleksey Yeschenko
>             Fix For: 3.x
>
>
> As part of CASSANDRA-9425 and a follow-up to CASSANDRA-9665, and a pre-requisite for
new schema change protocol, this ticket will do the following
> 1. Make the triggers {{HashMap}} immutable (new {{Triggers}} class)
> 2. Allow multiple 2i definitions per column in CFMetaData
> 3. <to be filled in>
> 4. Rename and move {{config.CFMetaData}} to {{schema.TableMetadata}}



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message