[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-10-07 Thread lantao yan (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13788592#comment-13788592
 ] 

lantao yan commented on CASSANDRA-5078:
---

Many thanks! So glad to see my contribution!

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Fix For: 2.0.2

 Attachments: 5078-v3.txt, 5078-v4.txt, 5078-v5.patch, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)



--
This message was sent by Atlassian JIRA
(v6.1#6144)


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-10-04 Thread lantao yan (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13785938#comment-13785938
 ] 

lantao yan commented on CASSANDRA-5078:
---

new patch submitted.

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Fix For: 2.0.2

 Attachments: 5078-v3.txt, 5078-v4.txt, 5078-v5.patch, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)



--
This message was sent by Atlassian JIRA
(v6.1#6144)


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-10-03 Thread lantao yan (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13785079#comment-13785079
 ] 

lantao yan commented on CASSANDRA-5078:
---

@Yuki Morishita, I checked that code, I think I was right about each thread 
will run against one column family store.
so two threads are working on the same column family with the same timestamp 
will never happen.

If given two compaction threads can work on the same column family 
store(directory, and files), we may have risk condition.
[~jbellis], correct me if I am wrong.

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: 5078-v3.txt, 5078-v4.txt, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)



--
This message was sent by Atlassian JIRA
(v6.1#6144)


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-10-03 Thread lantao yan (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13785336#comment-13785336
 ] 

lantao yan commented on CASSANDRA-5078:
---

[~yukim], [~jbellis], I checked that code, although two threads can work one 
the same column family. but in case of one thread is compacting, another thread 
would fail? I was worrying if two threads compacting the same set of files, we 
could have risk condition? given that, this problem is not actually changing 
the primary key.

ColumnFamilyStore.java
assert data.getCompacting().isEmpty() : data.getCompacting();

correct me if I am wrong. I will keep reading the code. thanks.

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: 5078-v3.txt, 5078-v4.txt, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)



--
This message was sent by Atlassian JIRA
(v6.1#6144)


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-10-03 Thread Yuki Morishita (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13785358#comment-13785358
 ] 

Yuki Morishita commented on CASSANDRA-5078:
---

Size-tiered/leveled compaction choose files that aren't compacting when 
determining candidates, and try marking acquired candidates as compacting. If 
marking fails, then compaction would be skipped(and go on to find next 
candidates).
So two or more threads can grab different set of SSTables inside the same 
ColumnFamily.

See, for example 
https://github.com/apache/cassandra/blob/cassandra-2.0/src/java/org/apache/cassandra/db/compaction/SizeTieredCompactionStrategy.java#L57
 and 
https://github.com/apache/cassandra/blob/cassandra-2.0/src/java/org/apache/cassandra/db/DataTracker.java#L188.

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: 5078-v3.txt, 5078-v4.txt, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)



--
This message was sent by Atlassian JIRA
(v6.1#6144)


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-10-03 Thread lantao yan (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13785524#comment-13785524
 ] 

lantao yan commented on CASSANDRA-5078:
---

Many thanks for the sharing! yes, I was wrong. Now I have better understanding. 
as you said, two threads can work on collections of sstables inside the same 
ks.cf.
I have no doubt now. One more question, is UUIDGen.getTimeUUID().toString() a 
correct way to generate the timeuuid as you suggest?

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: 5078-v3.txt, 5078-v4.txt, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)



--
This message was sent by Atlassian JIRA
(v6.1#6144)


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-10-03 Thread Jonathan Ellis (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13785529#comment-13785529
 ] 

Jonathan Ellis commented on CASSANDRA-5078:
---

Yes, but why turn it into a String?

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: 5078-v3.txt, 5078-v4.txt, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)



--
This message was sent by Atlassian JIRA
(v6.1#6144)


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-10-03 Thread lantao yan (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13785545#comment-13785545
 ] 

lantao yan commented on CASSANDRA-5078:
---

do you mean we can use UUID as column type?

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: 5078-v3.txt, 5078-v4.txt, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)



--
This message was sent by Atlassian JIRA
(v6.1#6144)


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-10-02 Thread Jonathan Ellis (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13784183#comment-13784183
 ] 

Jonathan Ellis commented on CASSANDRA-5078:
---

[~yanlantao], are you still working on this?

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: 5078-v3.txt, 5078-v4.txt, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)



--
This message was sent by Atlassian JIRA
(v6.1#6144)


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-10-02 Thread lantao yan (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13784239#comment-13784239
 ] 

lantao yan commented on CASSANDRA-5078:
---

sorry, i think i was wrong. I will read the code and consider Yuki's 
suggestion. it is still a little bit weird to me if two threads can work on the 
same column family.

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: 5078-v3.txt, 5078-v4.txt, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)



--
This message was sent by Atlassian JIRA
(v6.1#6144)


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-10-02 Thread Jonathan Ellis (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13784233#comment-13784233
 ] 

Jonathan Ellis commented on CASSANDRA-5078:
---

Hmm, if so this ticket did not get mentioned in a commit message or in CHANGES.

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: 5078-v3.txt, 5078-v4.txt, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)



--
This message was sent by Atlassian JIRA
(v6.1#6144)


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-09-11 Thread Yuki Morishita (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13764620#comment-13764620
 ] 

Yuki Morishita commented on CASSANDRA-5078:
---

bq. two threads are working on the same column family with the same timestamp?

Compactions can run in parallel. See CompactionExecutor in CompactionManager.
(Number of concurrency is set to number of cpu cores by default.)


 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: 5078-v3.txt, 5078-v4.txt, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-09-11 Thread lantao yan (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13764669#comment-13764669
 ] 

lantao yan commented on CASSANDRA-5078:
---

yeah, you are right. I thought it was a single thread executor.

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: 5078-v3.txt, 5078-v4.txt, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-09-10 Thread lantao yan (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13762966#comment-13762966
 ] 

lantao yan commented on CASSANDRA-5078:
---

sorry for the late update. pretty busy recently. even missed Jonathan's London 
meetup. :)
For the conflict on the entry, I don't quite get it. I read the code, it seems 
like one directory (one column family) will be owned by one thread? so there 
won't be any chances, two threads are working on the same column family with 
the same timestamp? I could be wrong. If I miss something, please let me know.

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: 5078-v3.txt, 5078-v4.txt, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-09-05 Thread Yuki Morishita (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13759336#comment-13759336
 ] 

Yuki Morishita commented on CASSANDRA-5078:
---

Thanks for the update.
One more thing that worries me is we have a chance to have conflict on the 
entry, since we can have same timestamp among compaction threads.
I suggest using just one timeuuid as primary key. Thoughts?

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: 5078-v3.txt, 5078-v4.txt, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-09-01 Thread lantao yan (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13755676#comment-13755676
 ] 

lantao yan commented on CASSANDRA-5078:
---

i attached a new patch. Please review it. Thanks.

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: 5078-v3.txt, 5078-v4.txt, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-31 Thread lantao yan (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13755563#comment-13755563
 ] 

lantao yan commented on CASSANDRA-5078:
---

I went through the feature 3974, and I may have some wrong solution for the ttl 
property I need to add. 
is the change by updating the cql to UPDATE system.transaction USING TTL 
DEFAULT_VALUE in class SystemKeyspace a correct fix?

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: 5078-v3.txt, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-31 Thread Jonathan Ellis (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13755652#comment-13755652
 ] 

Jonathan Ellis commented on CASSANDRA-5078:
---

CREATE TABLE ... WITH default_time_to_live=604800;

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: 5078-v3.txt, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-28 Thread Yuki Morishita (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13752524#comment-13752524
 ] 

Yuki Morishita commented on CASSANDRA-5078:
---

Couple of comments around JMX:

* Methods should return TabularData interface instead of TabularDataSupport
* I'd rather create class to hold CompositeType/CompositeData like streaming 
does(https://github.com/apache/cassandra/blob/cassandra-2.0/src/java/org/apache/cassandra/streaming/management/StreamStateCompositeData.java).
 That way, we don't need to construct type object every time we invoke.
* Cql rows to tabular data seems to be able to generalize for future use, but I 
don't think we need that at this time.

BTW, what is the step to remove unnecessary compaction history? Is `truncate 
system.compaction_history` the right way to do?

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: 5078-v3.txt, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-28 Thread Jonathan Ellis (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13752527#comment-13752527
 ] 

Jonathan Ellis commented on CASSANDRA-5078:
---

Should we also TTL it?

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: 5078-v3.txt, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-28 Thread lantao yan (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13752743#comment-13752743
 ] 

lantao yan commented on CASSANDRA-5078:
---

very good point for the second one. do i update it or somebody else will do it?

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: 5078-v3.txt, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-28 Thread Jonathan Ellis (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13752771#comment-13752771
 ] 

Jonathan Ellis commented on CASSANDRA-5078:
---

back to you :)

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: 5078-v3.txt, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-28 Thread lantao yan (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13752863#comment-13752863
 ] 

lantao yan commented on CASSANDRA-5078:
---

I have moved the code to a separate class to make sure type information only be 
construct once.
any example to ttl a system cf?

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: 5078-v3.txt, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-28 Thread Jonathan Ellis (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13752870#comment-13752870
 ] 

Jonathan Ellis commented on CASSANDRA-5078:
---

suggest setting a default ttl (CASSANDRA-3974) in cfmetadata creation

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: 5078-v3.txt, patch1.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-22 Thread lantao yan (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13747659#comment-13747659
 ] 

lantao yan commented on CASSANDRA-5078:
---

@Jonathan Ellis, do we save the compaction history of column families under 
system keyspace? or just customer spaces?

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Priority: Minor
  Labels: lhf
 Attachments: patch.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-22 Thread Jonathan Ellis (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13747780#comment-13747780
 ] 

Jonathan Ellis commented on CASSANDRA-5078:
---

We should probably exclude compaction_history itself, but including the other 
system tables is a good idea.

Other comments:
- Use CompositeData to expose this to JMX instead of formatted strings 
(https://github.com/apache/cassandra/tree/cassandra-2.0/src/java/org/apache/cassandra/streaming/management)
- Please follow the style guide at http://wiki.apache.org/cassandra/CodeStyle
- In general, don't extract blocks of code that are only called once into 
separate methods

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Priority: Minor
  Labels: lhf
 Attachments: patch.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-22 Thread Jonathan Ellis (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13747803#comment-13747803
 ] 

Jonathan Ellis commented on CASSANDRA-5078:
---

bq. Use CompositeData

Yuki suggests 
http://docs.oracle.com/javase/7/docs/api/javax/management/openmbean/TabularData.html
 may be a better fit.

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: patch.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-22 Thread lantao yan (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13747861#comment-13747861
 ] 

lantao yan commented on CASSANDRA-5078:
---

thanks for all the details. I will update the code this weekend. And I think I 
have found where the problem is. Two of three tests are running well now. I 
will keep reading the existing code base to make sure the change is reasonable.

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Assignee: lantao yan
Priority: Minor
  Labels: lhf
 Attachments: patch.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-21 Thread lantao yan (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13745895#comment-13745895
 ] 

lantao yan commented on CASSANDRA-5078:
---

i attached a new diff file between my branch and the latest trunk. the only 
change between the two diff files is ignoring all changes in cassandra.yaml 
file. btw, after applying the first diff file, what's the error?

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Priority: Minor
  Labels: lhf
 Attachments: diff.txt


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-21 Thread lantao yan (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13746551#comment-13746551
 ] 

lantao yan commented on CASSANDRA-5078:
---

is it because of the CompactionManager compacting compaction_history itself as 
well. and During this process, there is some issue?

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Priority: Minor
  Labels: lhf
 Attachments: patch.patch


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-20 Thread Jonathan Ellis (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13745077#comment-13745077
 ] 

Jonathan Ellis commented on CASSANDRA-5078:
---

Are you planning to take a stab at this, [~yanlantao]?

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Priority: Minor
  Labels: lhf

 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-20 Thread lantao yan (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13745093#comment-13745093
 ] 

lantao yan commented on CASSANDRA-5078:
---

yes. I have finish the coding to support the JMX and nodetool. everything looks 
fine, except when running three tests, the JVM crashes. 

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Priority: Minor
  Labels: lhf

 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-20 Thread Jonathan Ellis (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13745171#comment-13745171
 ] 

Jonathan Ellis commented on CASSANDRA-5078:
---

Are you running on Windows?  There are known problems w/ the test suite and tmp 
files there.

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Priority: Minor
  Labels: lhf

 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-20 Thread lantao yan (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13745271#comment-13745271
 ] 

lantao yan commented on CASSANDRA-5078:
---

unfortunately, I am using MAC. I run all tests on trunk, it is working fine; 
but after running against my code, JVM cashes. I am guessing, there is 
something wrong when some thread try to delete the temp file.

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Priority: Minor
  Labels: lhf

 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-20 Thread Brandon Williams (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13745280#comment-13745280
 ] 

Brandon Williams commented on CASSANDRA-5078:
-

which JVM and version are you using?  I would say just post the patch and we'll 
give it a try.

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Priority: Minor
  Labels: lhf

 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-20 Thread lantao yan (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13745381#comment-13745381
 ] 

lantao yan commented on CASSANDRA-5078:
---

i am using oracle JVM 7. How do I post the patch? push to a new branch?

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Priority: Minor
  Labels: lhf

 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-20 Thread Brandon Williams (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13745385#comment-13745385
 ] 

Brandon Williams commented on CASSANDRA-5078:
-

You can attach a diff in jira (more actions-attach files) or push a branch 
somewhere like your github, then post a comment with the link here.

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Priority: Minor
  Labels: lhf

 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-20 Thread lantao yan (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13745405#comment-13745405
 ] 

lantao yan commented on CASSANDRA-5078:
---

done. please check the attachment. one failing test is RecoveryManager3Test.

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Priority: Minor
  Labels: lhf
 Attachments: diff.txt


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-20 Thread Brandon Williams (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13745430#comment-13745430
 ] 

Brandon Williams commented on CASSANDRA-5078:
-

This diff doesn't apply cleanly for me.

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Priority: Minor
  Labels: lhf
 Attachments: diff.txt


 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-12 Thread lantao yan (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13737321#comment-13737321
 ] 

lantao yan commented on CASSANDRA-5078:
---

which system table should the compaction data save to? is there any sample code 
showing saving stuff to system table?

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Priority: Minor
  Labels: lhf

 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira


[jira] [Commented] (CASSANDRA-5078) save compaction merge counts in a system table

2013-08-12 Thread Jonathan Ellis (JIRA)

[ 
https://issues.apache.org/jira/browse/CASSANDRA-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13737330#comment-13737330
 ] 

Jonathan Ellis commented on CASSANDRA-5078:
---

I think you'd want a new table, something like

{code}
CREATE TABLE compaction_history (
text keyspace,
text columnfamily,
compacted_at timestamp,
bytes_in long,
bytes_out long,
rows_merged Mapint, long,
PRIMARY KEY (keyspace, columnfamily, compacted_at)
);
{code}

Look at CFMetadata for where the system tables are created, and SystemKeyspace 
for loading/storing examples.

 save compaction merge counts in a system table
 --

 Key: CASSANDRA-5078
 URL: https://issues.apache.org/jira/browse/CASSANDRA-5078
 Project: Cassandra
  Issue Type: Improvement
Reporter: Matthew F. Dennis
Priority: Minor
  Labels: lhf

 we should save the compaction merge stats from CASSANDRA-4894 in the system 
 table and probably expose them via JMX (and nodetool)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira