[jira] [Commented] (CASSANDRA-16735) Adding columns via ALTER TABLE can generate corrupt sstables
[
https://issues.apache.org/jira/browse/CASSANDRA-16735?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17428776#comment-17428776
]
Marcus Eriksson commented on CASSANDRA-16735:
-
we use {{schemaChangeInternal}} to introduce a schema mismatch between nodes,
and {{executeInternal}} to make sure we read repair later when doing a read
> Adding columns via ALTER TABLE can generate corrupt sstables
>
>
> Key: CASSANDRA-16735
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16735
> Project: Cassandra
> Issue Type: Bug
> Components: Consistency/Repair, Local/SSTable
>Reporter: Marcus Eriksson
>Assignee: Marcus Eriksson
>Priority: Normal
> Fix For: 3.0.25, 3.11.11, 4.0-rc2, 4.0
>
>
> This is similar to CASSANDRA-13004 and was caused by CASSANDRA-15899
> Basically the column placeholders introduced in 15899 can get read-repaired
> in to the memtable and later flushed to disk and in some cases this can
> conflict with the actual column (if the actual column is a collection for
> example) and cause CorruptSSTableExceptions.
> Fix is probably to just revert 15899, at least until if and when we find a
> solution that we can rely on. Will post that + test next week.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
-
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16735) Adding columns via ALTER TABLE can generate corrupt sstables
[ https://issues.apache.org/jira/browse/CASSANDRA-16735?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17428774#comment-17428774 ] Jacek Lewandowski commented on CASSANDRA-16735: --- Hi, I'm looking at `SchemaTest` modified in this ticket - what is that test doing? I mean - is the only way to reproduce that issue executing internal statements? > Adding columns via ALTER TABLE can generate corrupt sstables > > > Key: CASSANDRA-16735 > URL: https://issues.apache.org/jira/browse/CASSANDRA-16735 > Project: Cassandra > Issue Type: Bug > Components: Consistency/Repair, Local/SSTable >Reporter: Marcus Eriksson >Assignee: Marcus Eriksson >Priority: Normal > Fix For: 3.0.25, 3.11.11, 4.0-rc2, 4.0 > > > This is similar to CASSANDRA-13004 and was caused by CASSANDRA-15899 > Basically the column placeholders introduced in 15899 can get read-repaired > in to the memtable and later flushed to disk and in some cases this can > conflict with the actual column (if the actual column is a collection for > example) and cause CorruptSSTableExceptions. > Fix is probably to just revert 15899, at least until if and when we find a > solution that we can rely on. Will post that + test next week. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16735) Adding columns via ALTER TABLE can generate corrupt sstables
[ https://issues.apache.org/jira/browse/CASSANDRA-16735?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17370711#comment-17370711 ] Marcus Eriksson commented on CASSANDRA-16735: - thanks [~brandon.williams] https://issues.apache.org/jira/browse/CASSANDRA-16770 > Adding columns via ALTER TABLE can generate corrupt sstables > > > Key: CASSANDRA-16735 > URL: https://issues.apache.org/jira/browse/CASSANDRA-16735 > Project: Cassandra > Issue Type: Bug > Components: Consistency/Repair, Local/SSTable >Reporter: Marcus Eriksson >Assignee: Marcus Eriksson >Priority: Normal > Fix For: 3.0.25, 3.11.11, 4.0-rc2 > > > This is similar to CASSANDRA-13004 and was caused by CASSANDRA-15899 > Basically the column placeholders introduced in 15899 can get read-repaired > in to the memtable and later flushed to disk and in some cases this can > conflict with the actual column (if the actual column is a collection for > example) and cause CorruptSSTableExceptions. > Fix is probably to just revert 15899, at least until if and when we find a > solution that we can rely on. Will post that + test next week. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16735) Adding columns via ALTER TABLE can generate corrupt sstables
[ https://issues.apache.org/jira/browse/CASSANDRA-16735?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17369580#comment-17369580 ] Brandon Williams commented on CASSANDRA-16735: -- Note that this broke ~370 tests in the 3.0 to 3.11 upgrade: https://ci-cassandra.apache.org/job/Cassandra-3.11/186/ caused by our old friend "Unknown column cdc during deserialization" > Adding columns via ALTER TABLE can generate corrupt sstables > > > Key: CASSANDRA-16735 > URL: https://issues.apache.org/jira/browse/CASSANDRA-16735 > Project: Cassandra > Issue Type: Bug > Components: Consistency/Repair, Local/SSTable >Reporter: Marcus Eriksson >Assignee: Marcus Eriksson >Priority: Normal > Fix For: 3.0.25, 3.11.11, 4.0-rc2 > > > This is similar to CASSANDRA-13004 and was caused by CASSANDRA-15899 > Basically the column placeholders introduced in 15899 can get read-repaired > in to the memtable and later flushed to disk and in some cases this can > conflict with the actual column (if the actual column is a collection for > example) and cause CorruptSSTableExceptions. > Fix is probably to just revert 15899, at least until if and when we find a > solution that we can rely on. Will post that + test next week. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16735) Adding columns via ALTER TABLE can generate corrupt sstables
[ https://issues.apache.org/jira/browse/CASSANDRA-16735?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17366506#comment-17366506 ] Alex Petrov commented on CASSANDRA-16735: - LGTM +1 > Adding columns via ALTER TABLE can generate corrupt sstables > > > Key: CASSANDRA-16735 > URL: https://issues.apache.org/jira/browse/CASSANDRA-16735 > Project: Cassandra > Issue Type: Bug > Components: Consistency/Repair, Local/SSTable >Reporter: Marcus Eriksson >Assignee: Marcus Eriksson >Priority: Normal > Fix For: 4.0-rc2, 3.0.x, 3.11.x, 4.0-rc > > > This is similar to CASSANDRA-13004 and was caused by CASSANDRA-15899 > Basically the column placeholders introduced in 15899 can get read-repaired > in to the memtable and later flushed to disk and in some cases this can > conflict with the actual column (if the actual column is a collection for > example) and cause CorruptSSTableExceptions. > Fix is probably to just revert 15899, at least until if and when we find a > solution that we can rely on. Will post that + test next week. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16735) Adding columns via ALTER TABLE can generate corrupt sstables
[ https://issues.apache.org/jira/browse/CASSANDRA-16735?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17365536#comment-17365536 ] Ekaterina Dimitrova commented on CASSANDRA-16735: - Thank you for all your work [~marcuse]! :) I was wondering, should we really remove the _readWithSchemaDisagreement_ test? PS Indeed, sstableverify test is a known issue, not related to this patch > Adding columns via ALTER TABLE can generate corrupt sstables > > > Key: CASSANDRA-16735 > URL: https://issues.apache.org/jira/browse/CASSANDRA-16735 > Project: Cassandra > Issue Type: Bug > Components: Consistency/Repair, Local/SSTable >Reporter: Marcus Eriksson >Assignee: Marcus Eriksson >Priority: Normal > Fix For: 4.0-rc2, 3.0.x, 3.11.x, 4.0-rc > > > This is similar to CASSANDRA-13004 and was caused by CASSANDRA-15899 > Basically the column placeholders introduced in 15899 can get read-repaired > in to the memtable and later flushed to disk and in some cases this can > conflict with the actual column (if the actual column is a collection for > example) and cause CorruptSSTableExceptions. > Fix is probably to just revert 15899, at least until if and when we find a > solution that we can rely on. Will post that + test next week. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16735) Adding columns via ALTER TABLE can generate corrupt sstables
[ https://issues.apache.org/jira/browse/CASSANDRA-16735?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17365393#comment-17365393 ] Marcus Eriksson commented on CASSANDRA-16735: - removed the failing 15899 tests, now the cci runs look good (the sstableverify test fails locally without this patch as well) > Adding columns via ALTER TABLE can generate corrupt sstables > > > Key: CASSANDRA-16735 > URL: https://issues.apache.org/jira/browse/CASSANDRA-16735 > Project: Cassandra > Issue Type: Bug > Components: Consistency/Repair, Local/SSTable >Reporter: Marcus Eriksson >Assignee: Marcus Eriksson >Priority: Normal > Fix For: 4.0-rc2, 3.0.x, 3.11.x, 4.0-rc > > > This is similar to CASSANDRA-13004 and was caused by CASSANDRA-15899 > Basically the column placeholders introduced in 15899 can get read-repaired > in to the memtable and later flushed to disk and in some cases this can > conflict with the actual column (if the actual column is a collection for > example) and cause CorruptSSTableExceptions. > Fix is probably to just revert 15899, at least until if and when we find a > solution that we can rely on. Will post that + test next week. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16735) Adding columns via ALTER TABLE can generate corrupt sstables
[ https://issues.apache.org/jira/browse/CASSANDRA-16735?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17363162#comment-17363162 ] Brandon Williams commented on CASSANDRA-16735: -- Looks like this broke the in-jvm schema disagreement tests, and the 3.11 run didn't get the last commit to fix its build before it ran. > Adding columns via ALTER TABLE can generate corrupt sstables > > > Key: CASSANDRA-16735 > URL: https://issues.apache.org/jira/browse/CASSANDRA-16735 > Project: Cassandra > Issue Type: Bug > Components: Consistency/Repair, Local/SSTable >Reporter: Marcus Eriksson >Assignee: Marcus Eriksson >Priority: Normal > Fix For: 4.0-rc2, 3.0.x, 3.11.x, 4.0-rc > > > This is similar to CASSANDRA-13004 and was caused by CASSANDRA-15899 > Basically the column placeholders introduced in 15899 can get read-repaired > in to the memtable and later flushed to disk and in some cases this can > conflict with the actual column (if the actual column is a collection for > example) and cause CorruptSSTableExceptions. > Fix is probably to just revert 15899, at least until if and when we find a > solution that we can rely on. Will post that + test next week. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16735) Adding columns via ALTER TABLE can generate corrupt sstables
[ https://issues.apache.org/jira/browse/CASSANDRA-16735?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17361994#comment-17361994 ] Brandon Williams commented on CASSANDRA-16735: -- +1 to revert for 4.0.0 > Adding columns via ALTER TABLE can generate corrupt sstables > > > Key: CASSANDRA-16735 > URL: https://issues.apache.org/jira/browse/CASSANDRA-16735 > Project: Cassandra > Issue Type: Bug > Components: Consistency/Repair, Local/SSTable >Reporter: Marcus Eriksson >Assignee: Marcus Eriksson >Priority: Normal > Fix For: 4.0-rc2, 3.0.x, 3.11.x, 4.0-rc > > > This is similar to CASSANDRA-13004 and was caused by CASSANDRA-15899 > Basically the column placeholders introduced in 15899 can get read-repaired > in to the memtable and later flushed to disk and in some cases this can > conflict with the actual column (if the actual column is a collection for > example) and cause CorruptSSTableExceptions. > Fix is probably to just revert 15899, at least until if and when we find a > solution that we can rely on. Will post that + test next week. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-16735) Adding columns via ALTER TABLE can generate corrupt sstables
[ https://issues.apache.org/jira/browse/CASSANDRA-16735?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17361993#comment-17361993 ] Ekaterina Dimitrova commented on CASSANDRA-16735: - [~marcuse], I am not familiar with these works but I am trying to filter the blockers for 4.0.0 Jira board and I saw this ticket. Reading your thoughts in the description, do you think we can revert 15899 to unblock 4.0.0? > Adding columns via ALTER TABLE can generate corrupt sstables > > > Key: CASSANDRA-16735 > URL: https://issues.apache.org/jira/browse/CASSANDRA-16735 > Project: Cassandra > Issue Type: Bug > Components: Consistency/Repair, Local/SSTable >Reporter: Marcus Eriksson >Assignee: Marcus Eriksson >Priority: Normal > Fix For: 4.0-rc2, 3.0.x, 3.11.x, 4.0-rc > > > This is similar to CASSANDRA-13004 and was caused by CASSANDRA-15899 > Basically the column placeholders introduced in 15899 can get read-repaired > in to the memtable and later flushed to disk and in some cases this can > conflict with the actual column (if the actual column is a collection for > example) and cause CorruptSSTableExceptions. > Fix is probably to just revert 15899, at least until if and when we find a > solution that we can rely on. Will post that + test next week. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
