[
https://issues.apache.org/jira/browse/CASSANDRA-7193?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Ryan McGuire updated CASSANDRA-7193:
------------------------------------
Description:
I can create a table with the new query cache from CASSANDRA-5357:
{code}
CREATE TABLE status (user text, status_id timeuuid, status text, PRIMARY KEY
(user, status_id)) WITH caching = '{"keys":"ALL", "rows_per_partition":"10"}';
{code}
This method appears to work fine.
However, that is not the syntax mentioned in that ticket. It says to use a
rows_per_partition_to_cache setting instead, which does appear to work in cql:
{code}
CREATE TABLE status2 (user text, status_id timeuuid, status text, PRIMARY KEY
(user, status_id)) WITH rows_per_partition_to_cache = 200;
{code}
But that setting is not reflected in the table description, instead it still
shows NONE:
{code}
cqlsh:test> DESCRIBE TABLE status2;
CREATE TABLE test.status2 (
user text,
status_id timeuuid,
status text,
PRIMARY KEY (user, status_id)
) WITH CLUSTERING ORDER BY (status_id ASC)
AND caching = '{"keys":"ALL", "rows_per_partition":"NONE"}'
...
{code}
Similarly, alter table with that syntax, does not produce an error, but also
does not seem to affect the setting :
{code}
ALTER TABLE test.status WITH rows_per_partition_to_cache = 200;
{code}
was:
I can create a table with the new query cache from CASSANDRA-5357:
{code}
CREATE TABLE status (user text, status_id timeuuid, status text, PRIMARY KEY
(user, status_id)) WITH caching = '{"keys":"ALL", "rows_per_partition":"10"}';
{code}
This method appears to work fine.
However, that is not the syntax mentioned in that ticket. It says to use a
rows_per_partition_to_cache setting instead, which does appear to work in cql:
{code}
CREATE TABLE status2 (user text, status_id timeuuid, status text, PRIMARY KEY
(user, status_id)) WITH rows_per_partition_to_cache = 200;
{code}
But that setting is not reflected in the table description:
{code}
cqlsh:test> DESCRIBE TABLE status2;
CREATE TABLE test.status2 (
user text,
status_id timeuuid,
status text,
PRIMARY KEY (user, status_id)
) WITH CLUSTERING ORDER BY (status_id ASC)
AND caching = '{"keys":"ALL", "rows_per_partition":"NONE"}'
...
{code}
Similarly, alter table with that syntax, does not produce an error, but also
does not seem to affect the setting :
{code}
ALTER TABLE test.status WITH rows_per_partition_to_cache = 200;
{code}
> rows_per_partition_to_cache is not reflected in table DESCRIBE
> --------------------------------------------------------------
>
> Key: CASSANDRA-7193
> URL: https://issues.apache.org/jira/browse/CASSANDRA-7193
> Project: Cassandra
> Issue Type: Bug
> Reporter: Ryan McGuire
> Assignee: Marcus Eriksson
> Priority: Minor
>
> I can create a table with the new query cache from CASSANDRA-5357:
> {code}
> CREATE TABLE status (user text, status_id timeuuid, status text, PRIMARY KEY
> (user, status_id)) WITH caching = '{"keys":"ALL", "rows_per_partition":"10"}';
> {code}
> This method appears to work fine.
> However, that is not the syntax mentioned in that ticket. It says to use a
> rows_per_partition_to_cache setting instead, which does appear to work in cql:
> {code}
> CREATE TABLE status2 (user text, status_id timeuuid, status text, PRIMARY KEY
> (user, status_id)) WITH rows_per_partition_to_cache = 200;
> {code}
> But that setting is not reflected in the table description, instead it still
> shows NONE:
> {code}
> cqlsh:test> DESCRIBE TABLE status2;
> CREATE TABLE test.status2 (
> user text,
> status_id timeuuid,
> status text,
> PRIMARY KEY (user, status_id)
> ) WITH CLUSTERING ORDER BY (status_id ASC)
> AND caching = '{"keys":"ALL", "rows_per_partition":"NONE"}'
> ...
> {code}
> Similarly, alter table with that syntax, does not produce an error, but also
> does not seem to affect the setting :
> {code}
> ALTER TABLE test.status WITH rows_per_partition_to_cache = 200;
> {code}
--
This message was sent by Atlassian JIRA
(v6.2#6252)
