You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@cassandra.apache.org by "Jonathan Ellis (JIRA)" <ji...@apache.org> on 2009/04/03 22:52:12 UTC
[jira] Created: (CASSANDRA-52) Clean up SequenceFile.next to prep
for partitioning refactor
Clean up SequenceFile.next to prep for partitioning refactor
------------------------------------------------------------
Key: CASSANDRA-52
URL: https://issues.apache.org/jira/browse/CASSANDRA-52
Project: Cassandra
Issue Type: Improvement
Reporter: Jonathan Ellis
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (CASSANDRA-52) Clean up SequenceFile.next to prep
for partitioning refactor
Posted by "Jonathan Ellis (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-52?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12695961#action_12695961 ]
Jonathan Ellis commented on CASSANDRA-52:
-----------------------------------------
1. good point, I will add this tomorrow morning
2. I'm pretty sure this is because even though the test suite clears out SSTable files on disk, it can't clear out the other static stuff from StorageService, ColumnFamilyStore, et al. So sometimes the tests fail depending on what order they run in. If you run just the testTimeSort by itself it should pass reliably. (I have just done this in my IDE. Not sure how to do it in ant.)
> Clean up SequenceFile.next to prep for partitioning refactor
> ------------------------------------------------------------
>
> Key: CASSANDRA-52
> URL: https://issues.apache.org/jira/browse/CASSANDRA-52
> Project: Cassandra
> Issue Type: Improvement
> Reporter: Jonathan Ellis
> Attachments: 0001-CASSANDRA-52-write-get_columns_since-test-to-exercise.txt, 0002-reformat-code-mostly-whitespace-changes-to-keep-subs.txt, 0003-r-m-no-op-seeks.txt, 0004-refactor-SF.next-String-key-DataOutputBuffer-bufOut.txt, 0005-semantics-of-TimeFilter-include-all-columns-newer-tha.txt, 0006-emphasize-that-when-getting-a-time-based-slice-only-th.txt, 0007-merge-common-parts-of-next-timeRange-and-next-columnN.txt, 0008-combine-next-getData-overloads.txt, 0009-combine-overloads-of-SSTable.next-inline-getData.txt
>
>
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (CASSANDRA-52) Clean up SequenceFile.next to prep
for partitioning refactor
Posted by "Jonathan Ellis (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-52?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12695606#action_12695606 ]
Jonathan Ellis commented on CASSANDRA-52:
-----------------------------------------
9 combine overloads of SSTable.next; inline getData
8 combine next / getData overloads
7 merge common parts of next(timeRange) and next(columnNames)
6 emphasize that when getting a time-based slice only the CF name is used
5 semantics of TimeFilter (include all columns newer than some timestamp) mean we should use Long.MAX_VALUE as upper range, not current ms
4 refactor SF.next(String key, DataOutputBuffer bufOut, String cf, Coordinate section) to call its overload to reduce code duplication. this will make auditing the partitioner changes easier.
3 r/m no-op seeks
2 reformat code (mostly whitespace changes) to keep subsequent patches clean
1 CASSANDRA-52 write get_columns_since test to exercise sequenceFile.next(timeRange)
> Clean up SequenceFile.next to prep for partitioning refactor
> ------------------------------------------------------------
>
> Key: CASSANDRA-52
> URL: https://issues.apache.org/jira/browse/CASSANDRA-52
> Project: Cassandra
> Issue Type: Improvement
> Reporter: Jonathan Ellis
> Attachments: 0001-CASSANDRA-52-write-get_columns_since-test-to-exercise.txt, 0002-reformat-code-mostly-whitespace-changes-to-keep-subs.txt, 0003-r-m-no-op-seeks.txt, 0004-refactor-SF.next-String-key-DataOutputBuffer-bufOut.txt, 0005-semantics-of-TimeFilter-include-all-columns-newer-tha.txt, 0006-emphasize-that-when-getting-a-time-based-slice-only-th.txt, 0007-merge-common-parts-of-next-timeRange-and-next-columnN.txt, 0008-combine-next-getData-overloads.txt, 0009-combine-overloads-of-SSTable.next-inline-getData.txt
>
>
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (CASSANDRA-52) Clean up SequenceFile.next to prep
for partitioning refactor
Posted by "Jonathan Ellis (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-52?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Jonathan Ellis updated CASSANDRA-52:
------------------------------------
Attachment: 0009-combine-overloads-of-SSTable.next-inline-getData.txt
0008-combine-next-getData-overloads.txt
0007-merge-common-parts-of-next-timeRange-and-next-columnN.txt
0006-emphasize-that-when-getting-a-time-based-slice-only-th.txt
0005-semantics-of-TimeFilter-include-all-columns-newer-tha.txt
0004-refactor-SF.next-String-key-DataOutputBuffer-bufOut.txt
0003-r-m-no-op-seeks.txt
0002-reformat-code-mostly-whitespace-changes-to-keep-subs.txt
0001-CASSANDRA-52-write-get_columns_since-test-to-exercise.txt
> Clean up SequenceFile.next to prep for partitioning refactor
> ------------------------------------------------------------
>
> Key: CASSANDRA-52
> URL: https://issues.apache.org/jira/browse/CASSANDRA-52
> Project: Cassandra
> Issue Type: Improvement
> Reporter: Jonathan Ellis
> Attachments: 0001-CASSANDRA-52-write-get_columns_since-test-to-exercise.txt, 0002-reformat-code-mostly-whitespace-changes-to-keep-subs.txt, 0003-r-m-no-op-seeks.txt, 0004-refactor-SF.next-String-key-DataOutputBuffer-bufOut.txt, 0005-semantics-of-TimeFilter-include-all-columns-newer-tha.txt, 0006-emphasize-that-when-getting-a-time-based-slice-only-th.txt, 0007-merge-common-parts-of-next-timeRange-and-next-columnN.txt, 0008-combine-next-getData-overloads.txt, 0009-combine-overloads-of-SSTable.next-inline-getData.txt
>
>
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (CASSANDRA-52) Clean up SequenceFile.next to prep
for partitioning refactor
Posted by "Jun Rao (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-52?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12695960#action_12695960 ]
Jun Rao commented on CASSANDRA-52:
----------------------------------
I reviewed the patches. Overall, they look good. I got a couple of comments.
1. For tests like testTimeSort, it would be useful to add a test case where data is stored in both memtable and sstables.
2. ant test seems to fail on testTimeSort (after data is flushed to SSTable). However, when I stepped through the test in eclipse, it seemed to be ok. Not sure if others have the same issue.
> Clean up SequenceFile.next to prep for partitioning refactor
> ------------------------------------------------------------
>
> Key: CASSANDRA-52
> URL: https://issues.apache.org/jira/browse/CASSANDRA-52
> Project: Cassandra
> Issue Type: Improvement
> Reporter: Jonathan Ellis
> Attachments: 0001-CASSANDRA-52-write-get_columns_since-test-to-exercise.txt, 0002-reformat-code-mostly-whitespace-changes-to-keep-subs.txt, 0003-r-m-no-op-seeks.txt, 0004-refactor-SF.next-String-key-DataOutputBuffer-bufOut.txt, 0005-semantics-of-TimeFilter-include-all-columns-newer-tha.txt, 0006-emphasize-that-when-getting-a-time-based-slice-only-th.txt, 0007-merge-common-parts-of-next-timeRange-and-next-columnN.txt, 0008-combine-next-getData-overloads.txt, 0009-combine-overloads-of-SSTable.next-inline-getData.txt
>
>
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Resolved: (CASSANDRA-52) Clean up SequenceFile.next to prep
for partitioning refactor
Posted by "Jonathan Ellis (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-52?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Jonathan Ellis resolved CASSANDRA-52.
-------------------------------------
Resolution: Fixed
committed
> Clean up SequenceFile.next to prep for partitioning refactor
> ------------------------------------------------------------
>
> Key: CASSANDRA-52
> URL: https://issues.apache.org/jira/browse/CASSANDRA-52
> Project: Cassandra
> Issue Type: Improvement
> Reporter: Jonathan Ellis
> Attachments: 0001-CASSANDRA-52-write-get_columns_since-test-to-exercise.txt, 0002-reformat-code-mostly-whitespace-changes-to-keep-subs.txt, 0003-r-m-no-op-seeks.txt, 0004-refactor-SF.next-String-key-DataOutputBuffer-bufOut.txt, 0005-semantics-of-TimeFilter-include-all-columns-newer-tha.txt, 0006-emphasize-that-when-getting-a-time-based-slice-only-th.txt, 0007-merge-common-parts-of-next-timeRange-and-next-columnN.txt, 0008-combine-next-getData-overloads.txt, 0009-combine-overloads-of-SSTable.next-inline-getData.txt
>
>
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (CASSANDRA-52) Clean up SequenceFile.next to prep
for partitioning refactor
Posted by "Jun Rao (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-52?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12696205#action_12696205 ]
Jun Rao commented on CASSANDRA-52:
----------------------------------
This seems more like the issue pointed out by Johan https://issues.apache.org/jira/browse/CASSANDRA-57. When I ran just testTimeSort in eclipse, sometimes it succeeded and some other times it failed. I am wondering if this has to do with the synchronization of the flushing thread.
> Clean up SequenceFile.next to prep for partitioning refactor
> ------------------------------------------------------------
>
> Key: CASSANDRA-52
> URL: https://issues.apache.org/jira/browse/CASSANDRA-52
> Project: Cassandra
> Issue Type: Improvement
> Reporter: Jonathan Ellis
> Attachments: 0001-CASSANDRA-52-write-get_columns_since-test-to-exercise.txt, 0002-reformat-code-mostly-whitespace-changes-to-keep-subs.txt, 0003-r-m-no-op-seeks.txt, 0004-refactor-SF.next-String-key-DataOutputBuffer-bufOut.txt, 0005-semantics-of-TimeFilter-include-all-columns-newer-tha.txt, 0006-emphasize-that-when-getting-a-time-based-slice-only-th.txt, 0007-merge-common-parts-of-next-timeRange-and-next-columnN.txt, 0008-combine-next-getData-overloads.txt, 0009-combine-overloads-of-SSTable.next-inline-getData.txt
>
>
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Assigned: (CASSANDRA-52) Clean up SequenceFile.next to prep
for partitioning refactor
Posted by "Jonathan Ellis (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-52?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Jonathan Ellis reassigned CASSANDRA-52:
---------------------------------------
Assignee: Jonathan Ellis
> Clean up SequenceFile.next to prep for partitioning refactor
> ------------------------------------------------------------
>
> Key: CASSANDRA-52
> URL: https://issues.apache.org/jira/browse/CASSANDRA-52
> Project: Cassandra
> Issue Type: Improvement
> Reporter: Jonathan Ellis
> Assignee: Jonathan Ellis
> Attachments: 0001-CASSANDRA-52-write-get_columns_since-test-to-exercise.txt, 0002-reformat-code-mostly-whitespace-changes-to-keep-subs.txt, 0003-r-m-no-op-seeks.txt, 0004-refactor-SF.next-String-key-DataOutputBuffer-bufOut.txt, 0005-semantics-of-TimeFilter-include-all-columns-newer-tha.txt, 0006-emphasize-that-when-getting-a-time-based-slice-only-th.txt, 0007-merge-common-parts-of-next-timeRange-and-next-columnN.txt, 0008-combine-next-getData-overloads.txt, 0009-combine-overloads-of-SSTable.next-inline-getData.txt
>
>
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.