[
https://issues.apache.org/jira/browse/CASSANDRA-21451?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18088253#comment-18088253
]
guo Maxwell commented on CASSANDRA-21451:
-----------------------------------------
I think this is a nice ticket.
But I still want to ask , what is the differences between this issue with the
usage like this below:
(1) put all the cqls into a file name cql.txt
(2) use cqlsh -f cql.txt
if we just want to execute batch of cqls through some drivers ,this ticket may
be a good choice. But I think it is more important to design a standard batch
ddl grammar,
pg or mysql do not seem to have such grammar before.
> Support DDL batches
> -------------------
>
> Key: CASSANDRA-21451
> URL: https://issues.apache.org/jira/browse/CASSANDRA-21451
> Project: Apache Cassandra
> Issue Type: New Feature
> Reporter: Dmitry Konstantinov
> Priority: Normal
>
> If a schema is complicated and contains a lot of tables and UDTs then it can
> take a lot of time to apply DDLs one by one, especially on low level dev
> environments.
> To speedup the schema applying we can batch DDLs together, so we can amortize
> network interaction, flushing and transaction coordination time.
> Current logic and protocol for a schema change apply is already generic and
> ready for it.
> The change is only needed in CQL parsing and interpretation level.
> In case of TCM we can some DDLs in parallel but it is more complicated from a
> client side because it requires topological sorting or the changes.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
