[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15141014#comment-15141014
]
Sylvain Lebresne commented on CASSANDRA-7281:
-
lgtm, +1 (probably needs a minor rebasing though before committing).
> SELECT on tuple relations are broken for mixed ASC/DESC clustering order
>
>
> Key: CASSANDRA-7281
> URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
> Project: Cassandra
> Issue Type: Bug
>Reporter: Sylvain Lebresne
>Assignee: Marcin Szymaniuk
> Fix For: 3.4, 2.2.x, 3.0.x
>
> Attachments:
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v4.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v5.patch,
> 0001-Some-tests-for-CASSANDRA-7281.patch, 7281_unit_tests.txt
>
>
> As noted on
> [CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
> the tuple notation is broken when the clustering order mixes ASC and DESC
> directives because the range of data they describe don't correspond to a
> single continuous slice internally. To copy the example from CASSANDRA-6875:
> {noformat}
> cqlsh:ks> create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
> CLUSTERING ORDER BY (b DESC, c ASC);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
> cqlsh:ks> SELECT * FROM foo WHERE a=0;
> a | b | c
> ---+---+---
> 0 | 2 | 0
> 0 | 1 | 0
> 0 | 1 | 1
> 0 | 0 | 0
> (4 rows)
> cqlsh:ks> SELECT * FROM foo WHERE a=0 AND (b, c) > (1, 0);
> a | b | c
> ---+---+---
> 0 | 2 | 0
> (1 rows)
> {noformat}
> The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
> For that specific example we should generate 2 internal slices, but I believe
> that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15118917#comment-15118917
]
Benjamin Lerer commented on CASSANDRA-7281:
---
||patch||utests||dtests||
|[2.2|https://github.com/apache/cassandra/compare/trunk...blerer:7281-2.2]|[2.2|http://cassci.datastax.com/view/Dev/view/blerer/job/blerer-7281-2.2-testall/]|[2.2|http://cassci.datastax.com/view/Dev/view/blerer/job/blerer-7281-2.2-dtest/]|
|[3.0|https://github.com/apache/cassandra/compare/trunk...blerer:7281-3.0]|[3.0|http://cassci.datastax.com/view/Dev/view/blerer/job/blerer-7281-3.0-testall/]|[3.0|http://cassci.datastax.com/view/Dev/view/blerer/job/blerer-7281-3.0-dtest/]|
|[3.3|https://github.com/apache/cassandra/compare/trunk...blerer:7281-3.3]|[3.3|http://cassci.datastax.com/view/Dev/view/blerer/job/blerer-7281-3.3-testall/]|[3.3|http://cassci.datastax.com/view/Dev/view/blerer/job/blerer-7281-3.3-dtest/]|
|[trunk|https://github.com/apache/cassandra/compare/trunk...blerer:7281-trunk]|[trunk|http://cassci.datastax.com/view/Dev/view/blerer/job/blerer-7281-trunk-testall/]|[trunk|http://cassci.datastax.com/view/Dev/view/blerer/job/blerer-7281-trunk-dtest/]|
[~slebresne] could you review?
The patch for 2.2 is the same as before. It has just been rebased.
The patch for 3.0 is different for {{MultiCBuilder}},
{{PrimaryKeyRestrictionSet}} and {{PrimaryKeyRestrictionSetTest}} due to the
changes of CASSANDRA-8099.
The patch for 3.3 differs from the one of 3.0 for {{MultiCBuilder}} and
{{PrimaryKeyRestrictionSet}} due to the changes of CASSANDRA-10409 which
modified {{MultiCBuilder}}.
The patch for trunk is the same as the one of 3.3.
> SELECT on tuple relations are broken for mixed ASC/DESC clustering order
>
>
> Key: CASSANDRA-7281
> URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
> Project: Cassandra
> Issue Type: Bug
>Reporter: Sylvain Lebresne
>Assignee: Marcin Szymaniuk
> Fix For: 3.3, 2.2.x, 3.0.x
>
> Attachments:
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v4.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v5.patch,
> 0001-Some-tests-for-CASSANDRA-7281.patch, 7281_unit_tests.txt
>
>
> As noted on
> [CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
> the tuple notation is broken when the clustering order mixes ASC and DESC
> directives because the range of data they describe don't correspond to a
> single continuous slice internally. To copy the example from CASSANDRA-6875:
> {noformat}
> cqlsh:ks> create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
> CLUSTERING ORDER BY (b DESC, c ASC);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
> cqlsh:ks> SELECT * FROM foo WHERE a=0;
> a | b | c
> ---+---+---
> 0 | 2 | 0
> 0 | 1 | 0
> 0 | 1 | 1
> 0 | 0 | 0
> (4 rows)
> cqlsh:ks> SELECT * FROM foo WHERE a=0 AND (b, c) > (1, 0);
> a | b | c
> ---+---+---
> 0 | 2 | 0
> (1 rows)
> {noformat}
> The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
> For that specific example we should generate 2 internal slices, but I believe
> that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15115316#comment-15115316
]
Benjamin Lerer commented on CASSANDRA-7281:
---
Thanks for the work.
The code freeze for 3.3 will be next monday. By consequence, end of week will
probably be too late for the review process.
I will take it over.
> SELECT on tuple relations are broken for mixed ASC/DESC clustering order
>
>
> Key: CASSANDRA-7281
> URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
> Project: Cassandra
> Issue Type: Bug
>Reporter: Sylvain Lebresne
>Assignee: Marcin Szymaniuk
> Fix For: 3.3, 2.2.x, 3.0.x
>
> Attachments:
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v4.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v5.patch,
> 0001-Some-tests-for-CASSANDRA-7281.patch, 7281_unit_tests.txt
>
>
> As noted on
> [CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
> the tuple notation is broken when the clustering order mixes ASC and DESC
> directives because the range of data they describe don't correspond to a
> single continuous slice internally. To copy the example from CASSANDRA-6875:
> {noformat}
> cqlsh:ks> create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
> CLUSTERING ORDER BY (b DESC, c ASC);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
> cqlsh:ks> SELECT * FROM foo WHERE a=0;
> a | b | c
> ---+---+---
> 0 | 2 | 0
> 0 | 1 | 0
> 0 | 1 | 1
> 0 | 0 | 0
> (4 rows)
> cqlsh:ks> SELECT * FROM foo WHERE a=0 AND (b, c) > (1, 0);
> a | b | c
> ---+---+---
> 0 | 2 | 0
> (1 rows)
> {noformat}
> The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
> For that specific example we should generate 2 internal slices, but I believe
> that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15114522#comment-15114522
]
Marcin Szymaniuk commented on CASSANDRA-7281:
-
To be honest I underestimated time I need to get familiar with 3.x. There are
some significant changes around bounds. So far I just rewritten some test
Benjamin added. Depends on your schedule, I will have a look at it later this
week/weekend but I understand it might be too late for your release. Please let
me know if you decide you start working on this on your own so I don't do it in
parallel. I'm attaching two tests I rewritten - It can save you some typing.
> SELECT on tuple relations are broken for mixed ASC/DESC clustering order
>
>
> Key: CASSANDRA-7281
> URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
> Project: Cassandra
> Issue Type: Bug
>Reporter: Sylvain Lebresne
>Assignee: Marcin Szymaniuk
> Fix For: 3.3, 2.2.x, 3.0.x
>
> Attachments:
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v4.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v5.patch,
> 7281_unit_tests.txt
>
>
> As noted on
> [CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
> the tuple notation is broken when the clustering order mixes ASC and DESC
> directives because the range of data they describe don't correspond to a
> single continuous slice internally. To copy the example from CASSANDRA-6875:
> {noformat}
> cqlsh:ks> create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
> CLUSTERING ORDER BY (b DESC, c ASC);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
> cqlsh:ks> SELECT * FROM foo WHERE a=0;
> a | b | c
> ---+---+---
> 0 | 2 | 0
> 0 | 1 | 0
> 0 | 1 | 1
> 0 | 0 | 0
> (4 rows)
> cqlsh:ks> SELECT * FROM foo WHERE a=0 AND (b, c) > (1, 0);
> a | b | c
> ---+---+---
> 0 | 2 | 0
> (1 rows)
> {noformat}
> The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
> For that specific example we should generate 2 internal slices, but I believe
> that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15108830#comment-15108830
]
Benjamin Lerer commented on CASSANDRA-7281:
---
It would be perfect. :-)
> SELECT on tuple relations are broken for mixed ASC/DESC clustering order
>
>
> Key: CASSANDRA-7281
> URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
> Project: Cassandra
> Issue Type: Bug
>Reporter: Sylvain Lebresne
>Assignee: Marcin Szymaniuk
> Fix For: 3.3, 2.2.x, 3.0.x
>
> Attachments:
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v4.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v5.patch,
> 7281_unit_tests.txt
>
>
> As noted on
> [CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
> the tuple notation is broken when the clustering order mixes ASC and DESC
> directives because the range of data they describe don't correspond to a
> single continuous slice internally. To copy the example from CASSANDRA-6875:
> {noformat}
> cqlsh:ks> create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
> CLUSTERING ORDER BY (b DESC, c ASC);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
> cqlsh:ks> SELECT * FROM foo WHERE a=0;
> a | b | c
> ---+---+---
> 0 | 2 | 0
> 0 | 1 | 0
> 0 | 1 | 1
> 0 | 0 | 0
> (4 rows)
> cqlsh:ks> SELECT * FROM foo WHERE a=0 AND (b, c) > (1, 0);
> a | b | c
> ---+---+---
> 0 | 2 | 0
> (1 rows)
> {noformat}
> The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
> For that specific example we should generate 2 internal slices, but I believe
> that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15108230#comment-15108230
]
Marcin Szymaniuk commented on CASSANDRA-7281:
-
I will do it over the weekend if it's not too late for you.
> SELECT on tuple relations are broken for mixed ASC/DESC clustering order
>
>
> Key: CASSANDRA-7281
> URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
> Project: Cassandra
> Issue Type: Bug
>Reporter: Sylvain Lebresne
>Assignee: Marcin Szymaniuk
> Fix For: 3.3, 2.2.x, 3.0.x
>
> Attachments:
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v4.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v5.patch,
> 7281_unit_tests.txt
>
>
> As noted on
> [CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
> the tuple notation is broken when the clustering order mixes ASC and DESC
> directives because the range of data they describe don't correspond to a
> single continuous slice internally. To copy the example from CASSANDRA-6875:
> {noformat}
> cqlsh:ks> create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
> CLUSTERING ORDER BY (b DESC, c ASC);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
> cqlsh:ks> SELECT * FROM foo WHERE a=0;
> a | b | c
> ---+---+---
> 0 | 2 | 0
> 0 | 1 | 0
> 0 | 1 | 1
> 0 | 0 | 0
> (4 rows)
> cqlsh:ks> SELECT * FROM foo WHERE a=0 AND (b, c) > (1, 0);
> a | b | c
> ---+---+---
> 0 | 2 | 0
> (1 rows)
> {noformat}
> The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
> For that specific example we should generate 2 internal slices, but I believe
> that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15095858#comment-15095858
]
Benjamin Lerer commented on CASSANDRA-7281:
---
[~marcinszymaniuk] do you have time to make the patches for 3.0 and 3.3 in the
coming days? We would like to have this ticket fixed for the 3.3 release.
> SELECT on tuple relations are broken for mixed ASC/DESC clustering order
>
>
> Key: CASSANDRA-7281
> URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
> Project: Cassandra
> Issue Type: Bug
>Reporter: Sylvain Lebresne
>Assignee: Marcin Szymaniuk
> Fix For: 2.1.x
>
> Attachments:
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v4.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v5.patch,
> 7281_unit_tests.txt
>
>
> As noted on
> [CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
> the tuple notation is broken when the clustering order mixes ASC and DESC
> directives because the range of data they describe don't correspond to a
> single continuous slice internally. To copy the example from CASSANDRA-6875:
> {noformat}
> cqlsh:ks> create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
> CLUSTERING ORDER BY (b DESC, c ASC);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
> cqlsh:ks> SELECT * FROM foo WHERE a=0;
> a | b | c
> ---+---+---
> 0 | 2 | 0
> 0 | 1 | 0
> 0 | 1 | 1
> 0 | 0 | 0
> (4 rows)
> cqlsh:ks> SELECT * FROM foo WHERE a=0 AND (b, c) > (1, 0);
> a | b | c
> ---+---+---
> 0 | 2 | 0
> (1 rows)
> {noformat}
> The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
> For that specific example we should generate 2 internal slices, but I believe
> that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15094027#comment-15094027
]
Sylvain Lebresne commented on CASSANDRA-7281:
-
Had a look at [~blerer]'s modified version and it looks good, but we obviously
need patches for 3.0 and upwards (as well as links to the CI results). I don't
think we'll want to commit this to 2.1 at this point however, so I suggest not
wasting time on that.
> SELECT on tuple relations are broken for mixed ASC/DESC clustering order
>
>
> Key: CASSANDRA-7281
> URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
> Project: Cassandra
> Issue Type: Bug
>Reporter: Sylvain Lebresne
>Assignee: Marcin Szymaniuk
> Fix For: 2.1.x
>
> Attachments:
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v4.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v5.patch,
> 7281_unit_tests.txt
>
>
> As noted on
> [CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
> the tuple notation is broken when the clustering order mixes ASC and DESC
> directives because the range of data they describe don't correspond to a
> single continuous slice internally. To copy the example from CASSANDRA-6875:
> {noformat}
> cqlsh:ks> create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
> CLUSTERING ORDER BY (b DESC, c ASC);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
> cqlsh:ks> SELECT * FROM foo WHERE a=0;
> a | b | c
> ---+---+---
> 0 | 2 | 0
> 0 | 1 | 0
> 0 | 1 | 1
> 0 | 0 | 0
> (4 rows)
> cqlsh:ks> SELECT * FROM foo WHERE a=0 AND (b, c) > (1, 0);
> a | b | c
> ---+---+---
> 0 | 2 | 0
> (1 rows)
> {noformat}
> The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
> For that specific example we should generate 2 internal slices, but I believe
> that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15034102#comment-15034102
]
Marcin Szymaniuk commented on CASSANDRA-7281:
-
I'm still interested. I will have a look at your patch.
> SELECT on tuple relations are broken for mixed ASC/DESC clustering order
>
>
> Key: CASSANDRA-7281
> URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
> Project: Cassandra
> Issue Type: Bug
>Reporter: Sylvain Lebresne
>Assignee: Marcin Szymaniuk
> Fix For: 2.1.x
>
> Attachments:
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v4.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v5.patch,
> 7281_unit_tests.txt
>
>
> As noted on
> [CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
> the tuple notation is broken when the clustering order mixes ASC and DESC
> directives because the range of data they describe don't correspond to a
> single continuous slice internally. To copy the example from CASSANDRA-6875:
> {noformat}
> cqlsh:ks> create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
> CLUSTERING ORDER BY (b DESC, c ASC);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
> cqlsh:ks> SELECT * FROM foo WHERE a=0;
> a | b | c
> ---+---+---
> 0 | 2 | 0
> 0 | 1 | 0
> 0 | 1 | 1
> 0 | 0 | 0
> (4 rows)
> cqlsh:ks> SELECT * FROM foo WHERE a=0 AND (b, c) > (1, 0);
> a | b | c
> ---+---+---
> 0 | 2 | 0
> (1 rows)
> {noformat}
> The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
> For that specific example we should generate 2 internal slices, but I believe
> that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15027102#comment-15027102
]
Benjamin Lerer commented on CASSANDRA-7281:
---
[~marcinszymaniuk] do you have any comments on my modified version?
Are you still interested in providing a patch for 2.1 and 3.0?
> SELECT on tuple relations are broken for mixed ASC/DESC clustering order
>
>
> Key: CASSANDRA-7281
> URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
> Project: Cassandra
> Issue Type: Bug
>Reporter: Sylvain Lebresne
>Assignee: Marcin Szymaniuk
> Fix For: 2.1.x
>
> Attachments:
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v4.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v5.patch,
> 7281_unit_tests.txt
>
>
> As noted on
> [CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
> the tuple notation is broken when the clustering order mixes ASC and DESC
> directives because the range of data they describe don't correspond to a
> single continuous slice internally. To copy the example from CASSANDRA-6875:
> {noformat}
> cqlsh:ks> create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
> CLUSTERING ORDER BY (b DESC, c ASC);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
> cqlsh:ks> SELECT * FROM foo WHERE a=0;
> a | b | c
> ---+---+---
> 0 | 2 | 0
> 0 | 1 | 0
> 0 | 1 | 1
> 0 | 0 | 0
> (4 rows)
> cqlsh:ks> SELECT * FROM foo WHERE a=0 AND (b, c) > (1, 0);
> a | b | c
> ---+---+---
> 0 | 2 | 0
> (1 rows)
> {noformat}
> The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
> For that specific example we should generate 2 internal slices, but I believe
> that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=14966879#comment-14966879
]
Benjamin Lerer commented on CASSANDRA-7281:
---
Sorry for not reviewing the patch sooner.
Your patch suffer from the same problem as I mentionned earlier. It will not
work with a mix of single and multi column restrictions like:
{{SELECT * FROM myTable WHERE partitionKey = 1 AND clustering_1 IN (1, 2) AND
clustering_2 = 0 AND (clustering_3, clustering_4) > (2, 3);}}
Another problem is that a big part of the code in {{PrimaryKeyRestrictionSet}}
is relative to {{MultiColumnRestrictions}} and should be in my opinion
encapsulate within it. Having this code within {{PrimaryKeyRestrictionSet}}
will make the implementation of an issue like CASSANDRA-8855 much more
complicated.
I have pushed modified version of the patch for 2.2
[here|https://github.com/apache/cassandra/compare/trunk...blerer:7281-2.2] (it
runs fine on CI).
Feel free to provide some feedbacks on it.
> SELECT on tuple relations are broken for mixed ASC/DESC clustering order
>
>
> Key: CASSANDRA-7281
> URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
> Project: Cassandra
> Issue Type: Bug
>Reporter: Sylvain Lebresne
>Assignee: Marcin Szymaniuk
> Fix For: 2.1.x
>
> Attachments:
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v4.patch,
> 0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v5.patch,
> 7281_unit_tests.txt
>
>
> As noted on
> [CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
> the tuple notation is broken when the clustering order mixes ASC and DESC
> directives because the range of data they describe don't correspond to a
> single continuous slice internally. To copy the example from CASSANDRA-6875:
> {noformat}
> cqlsh:ks> create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
> CLUSTERING ORDER BY (b DESC, c ASC);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
> cqlsh:ks> INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
> cqlsh:ks> SELECT * FROM foo WHERE a=0;
> a | b | c
> ---+---+---
> 0 | 2 | 0
> 0 | 1 | 0
> 0 | 1 | 1
> 0 | 0 | 0
> (4 rows)
> cqlsh:ks> SELECT * FROM foo WHERE a=0 AND (b, c) > (1, 0);
> a | b | c
> ---+---+---
> 0 | 2 | 0
> (1 rows)
> {noformat}
> The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
> For that specific example we should generate 2 internal slices, but I believe
> that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14586621#comment-14586621
]
Marcin Szymaniuk commented on CASSANDRA-7281:
-
All right. I will merge it with unit tests I already have (It might be that
those from dtest are not necessary anymore).
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Assignee: Marcin Szymaniuk
Fix For: 2.1.x
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v4.patch,
7281_unit_tests.txt
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14587132#comment-14587132
]
Stefania commented on CASSANDRA-7281:
-
Thanks!
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Assignee: Marcin Szymaniuk
Fix For: 2.1.x
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v4.patch,
7281_unit_tests.txt
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14568783#comment-14568783
]
Benjamin Lerer commented on CASSANDRA-7281:
---
Sorry, my comment was really not that clear.
In 2.1: I think that your code should tackle the problem within the
{{SelectStatement.buildBound}} method. The tests for that method are within
{{SelectStatementTest}}.
In 2.2 you are right the code should be within
{{MultiColumnRestriction.Slice.appendBoundTo}} method and the tests are within
{{PrimaryKeyRestrictionSetTest }}.
I expect that if you manage to fix the problem in 2.1 as part of the
{{SelectStatement.buildBound}} method you should be able to make it work within
2.2 without too much difficulties by extracting your fix and putting it within
{{MultiColumnRestriction.Slice.appendBoundTo}}.
Hope it helps.
Do not hesitate to contact me on IRC or by email if you have more questions
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Assignee: Marcin Szymaniuk
Fix For: 2.1.x
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v4.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14568757#comment-14568757
]
Marcin Szymaniuk commented on CASSANDRA-7281:
-
In 2.2 I cannot see classes/methods mentioned in your comment.
From quick look it seems to me that:
- MultiColumnRestriction.Slice#appendBoundTo is the place to implement the logic
- PrimaryKeyRestrictionSetTest is good place to create more of unit-style tests.
Can you please have a look and confirm?
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Assignee: Marcin Szymaniuk
Fix For: 2.1.x
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v4.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14561778#comment-14561778
]
Marcin Szymaniuk commented on CASSANDRA-7281:
-
Ok I will have a look. I guess we should change Fix Version/s then?
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Assignee: Marcin Szymaniuk
Fix For: 2.1.x
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v4.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14561779#comment-14561779
]
Marcin Szymaniuk commented on CASSANDRA-7281:
-
Ok I will have a look. I guess we should change Fix Version/s then?
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Assignee: Marcin Szymaniuk
Fix For: 2.1.x
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v4.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14558870#comment-14558870
]
Benjamin Lerer commented on CASSANDRA-7281:
---
Thanks for the patch. It is nice that you added all those unit tests.
I had a quick look at the patch and I have the following comments:
# CASSANDRA-8613 has changed the way things are working. You can now have a mix
a tulpe relations and non tulpe relations. This means that you first relation
might not be a tulpe relation or a slice tulpe relation.
# In 2.2 {{SelectStatement}} has been refactored to simplify the code and allow
us to implements more complex features. Your approach of creating new
restrictions will not fit properly there.
I believe that you should be able to skip this restrictions creation step and
implements the logic into {{buildBound}} by going directly from the original
restriction to the composites.
FYI: The {{buildBond}} unit tests are located in {{SelectStatementTest}}.
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Assignee: Marcin Szymaniuk
Fix For: 2.1.x
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v4.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14558125#comment-14558125
]
Marcin Szymaniuk commented on CASSANDRA-7281:
-
I Uploaded a patch yesterday in case you don't get notifications for that.
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Assignee: Marcin Szymaniuk
Fix For: 2.1.x
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v4.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14558124#comment-14558124
]
Marcin Szymaniuk commented on CASSANDRA-7281:
-
I Uploaded a patch yesterday in case you don't get notifications for that.
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Assignee: Marcin Szymaniuk
Fix For: 2.1.x
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v4.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14550010#comment-14550010
]
Marcin Szymaniuk commented on CASSANDRA-7281:
-
I will provide another patch by the end of this week
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Assignee: Marcin Szymaniuk
Fix For: 2.1.x
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14550027#comment-14550027
]
Benjamin Lerer commented on CASSANDRA-7281:
---
[~marcinszymaniuk] Great! Thanks
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Assignee: Marcin Szymaniuk
Fix For: 2.1.x
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14539576#comment-14539576
]
Benjamin Lerer commented on CASSANDRA-7281:
---
[~marcinszymaniuk] Do you still want to provide a patch or should I reasign the
issue? The issue has been open for nearly a year and I would like to get rid of
it as it is one of the last issues that we have on tuple relations.
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Assignee: Marcin Szymaniuk
Fix For: 2.1.x
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14352704#comment-14352704
]
Marcin Szymaniuk commented on CASSANDRA-7281:
-
Ok, cool
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Assignee: Marcin Szymaniuk
Fix For: 2.1.4
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14350102#comment-14350102
]
Benjamin Lerer commented on CASSANDRA-7281:
---
[~marcinszymaniuk] CASSANDRA-8613 as been committed.
You are good to go :-)
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Assignee: Marcin Szymaniuk
Fix For: 2.1.4
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14300027#comment-14300027
]
Marcin Szymaniuk commented on CASSANDRA-7281:
-
Sure, I started watching your ticket.
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Assignee: Marcin Szymaniuk
Fix For: 2.1.3, 2.0.13
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14298609#comment-14298609
]
Benjamin Lerer commented on CASSANDRA-7281:
---
[~marcinszymaniuk] I am currently working on CASSANDRA-8613 and the changes
requireded for it will impact a lot the code of {{SelectStatement}} used to
build the requestedBounds.
If you can, it might be better to wait a bit before writing your patch.
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Assignee: Marcin Szymaniuk
Fix For: 2.1.3, 2.0.13
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14294121#comment-14294121
]
Benjamin Lerer commented on CASSANDRA-7281:
---
[~marcinszymaniuk] do you want to provide another patch or should I reassign
the ticket?
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Assignee: Marcin Szymaniuk
Fix For: 2.1.3, 2.0.13
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14294313#comment-14294313
]
Marcin Szymaniuk commented on CASSANDRA-7281:
-
Ah yes I will provide another one. I missed your message.
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Assignee: Marcin Szymaniuk
Fix For: 2.1.3, 2.0.13
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14276925#comment-14276925
]
Benjamin Lerer commented on CASSANDRA-7281:
---
* It seems that the patch does not work when the two sides of the slice are
specified. Queries like:
{code}
SELECT * FROM foo WHERE a=0 AND (b) (0) AND (b) = (1);
{code}
or
{code}
SELECT * FROM %s WHERE a=0 AND (b, c) (1, 0) AND (b, c) (2, 0);
{code}
do not returns the good rows.
* It would be good to also add some unit tests to {{MultiColumnRelationTest}}
to cover all of the possible cases
* The code is quite heavy and will not merge in 3.0 due to the
{{SelectStatement}} refactoring.You should be able to fix this ticket by
changing only the {{SelectStatement.buildMultiColumnSliceBound}}
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Assignee: Marcin Szymaniuk
Fix For: 2.1.3, 2.0.13
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14250021#comment-14250021
]
Philip Thompson commented on CASSANDRA-7281:
Any chance this could get reviewed before the holidays?
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Assignee: Marcin Szymaniuk
Fix For: 2.0.12
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14250245#comment-14250245
]
Benjamin Lerer commented on CASSANDRA-7281:
---
I will try but the chances are low.
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Assignee: Marcin Szymaniuk
Fix For: 2.0.12
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14250263#comment-14250263
]
Philip Thompson commented on CASSANDRA-7281:
Okay, that is fine. Sylvain mentioned a possible 2.1.3 release by the end of
the calendar year. If this patch makes it into that release, that is soon
enough.
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Assignee: Marcin Szymaniuk
Fix For: 2.0.12, 2.1.3
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14227974#comment-14227974
]
Marcin Szymaniuk commented on CASSANDRA-7281:
-
Ah sorry. Just attached patch for 2.0.
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Fix For: 2.0.12
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v3.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14224632#comment-14224632
]
Benjamin Lerer commented on CASSANDRA-7281:
---
The ticket is open for the version 2.0 but it seems that the patch has been
built for an other version. Marcin, could you backport your fix for version 2.0.
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Fix For: 2.0.12
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch,
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-v2.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14215488#comment-14215488
]
Michael Shuler commented on CASSANDRA-7281:
---
Nit: if you're going to remove lines in
{{src/java/org/apache/cassandra/db/filter/ColumnSlice.java}}, remove them,
please, don't comment them out.
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Fix For: 2.0.12
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (CASSANDRA-7281) SELECT on tuple relations are broken for mixed ASC/DESC clustering order
[
https://issues.apache.org/jira/browse/CASSANDRA-7281?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14215496#comment-14215496
]
Michael Shuler commented on CASSANDRA-7281:
---
Missed the dtest comment - send those as a pull request on
https://github.com/riptano/cassandra-dtest - thanks!
SELECT on tuple relations are broken for mixed ASC/DESC clustering order
Key: CASSANDRA-7281
URL: https://issues.apache.org/jira/browse/CASSANDRA-7281
Project: Cassandra
Issue Type: Bug
Reporter: Sylvain Lebresne
Fix For: 2.0.12
Attachments:
0001-CASSANDRA-7281-SELECT-on-tuple-relations-are-broken-.patch
As noted on
[CASSANDRA-6875|https://issues.apache.org/jira/browse/CASSANDRA-6875?focusedCommentId=13992153page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13992153],
the tuple notation is broken when the clustering order mixes ASC and DESC
directives because the range of data they describe don't correspond to a
single continuous slice internally. To copy the example from CASSANDRA-6875:
{noformat}
cqlsh:ks create table foo (a int, b int, c int, PRIMARY KEY (a, b, c)) WITH
CLUSTERING ORDER BY (b DESC, c ASC);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 2, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 0);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 1, 1);
cqlsh:ks INSERT INTO foo (a, b, c) VALUES (0, 0, 0);
cqlsh:ks SELECT * FROM foo WHERE a=0;
a | b | c
---+---+---
0 | 2 | 0
0 | 1 | 0
0 | 1 | 1
0 | 0 | 0
(4 rows)
cqlsh:ks SELECT * FROM foo WHERE a=0 AND (b, c) (1, 0);
a | b | c
---+---+---
0 | 2 | 0
(1 rows)
{noformat}
The last query should really return {{(0, 2, 0)}} and {{(0, 1, 1)}}.
For that specific example we should generate 2 internal slices, but I believe
that with more clustering columns we may have more slices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
