[jira] [Commented] (SOLR-3080) We should consider removing shard info from Zk when you explicitly unload a SolrCore.

2012-02-26 Thread Yonik Seeley (Commented) (JIRA)

[ 
https://issues.apache.org/jira/browse/SOLR-3080?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13216842#comment-13216842
 ] 

Yonik Seeley commented on SOLR-3080:


After removing isLeader==true test for tlog recovery, I ran for 4 hours w/o a 
fail with deletes turned off.  Turning those on now...

 We should consider removing shard info from Zk when you explicitly unload a 
 SolrCore.
 -

 Key: SOLR-3080
 URL: https://issues.apache.org/jira/browse/SOLR-3080
 Project: Solr
  Issue Type: Improvement
  Components: SolrCloud
Reporter: Mark Miller
Assignee: Sami Siren
Priority: Minor
 Fix For: 4.0

 Attachments: SOLR-3080.patch, SOLR-3080.patch, 
 SOLR-3080_smaller_sync.patch




--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: 
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira



-
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org



[jira] [Commented] (SOLR-3080) We should consider removing shard info from Zk when you explicitly unload a SolrCore.

2012-02-25 Thread Yonik Seeley (Commented) (JIRA)

[ 
https://issues.apache.org/jira/browse/SOLR-3080?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13216633#comment-13216633
 ] 

Yonik Seeley commented on SOLR-3080:


OK, here's what I think is happening around the ChaosMonkeySafeLeaderTest 
failures I'm seeing.

The missing docs actually appear in the startingVersions (i.e. they are in 
the transaction log).
We peersync and don't request those versions because we already have them.
*But* I don't think we ever called UpdateLog.recoverFromLog() to actually apply 
them to the index!

In ZkController.register() recoverFromLog is only called if isLeader==true.  
Perhaps this is a holdover from when peersync was not done at startup and hence 
it didn't make sense to try and recover from your logs when you were just going 
to replicate the leader index anyway?

 We should consider removing shard info from Zk when you explicitly unload a 
 SolrCore.
 -

 Key: SOLR-3080
 URL: https://issues.apache.org/jira/browse/SOLR-3080
 Project: Solr
  Issue Type: Improvement
  Components: SolrCloud
Reporter: Mark Miller
Assignee: Sami Siren
Priority: Minor
 Fix For: 4.0

 Attachments: SOLR-3080.patch, SOLR-3080.patch, 
 SOLR-3080_smaller_sync.patch




--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: 
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira



-
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org



[jira] [Commented] (SOLR-3080) We should consider removing shard info from Zk when you explicitly unload a SolrCore.

2012-02-24 Thread Sami Siren (Commented) (JIRA)

[ 
https://issues.apache.org/jira/browse/SOLR-3080?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13215760#comment-13215760
 ] 

Sami Siren commented on SOLR-3080:
--

+1, nice!

 We should consider removing shard info from Zk when you explicitly unload a 
 SolrCore.
 -

 Key: SOLR-3080
 URL: https://issues.apache.org/jira/browse/SOLR-3080
 Project: Solr
  Issue Type: Improvement
  Components: SolrCloud
Reporter: Mark Miller
Assignee: Sami Siren
Priority: Minor
 Fix For: 4.0

 Attachments: SOLR-3080.patch, SOLR-3080.patch, 
 SOLR-3080_smaller_sync.patch




--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: 
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira



-
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org



[jira] [Commented] (SOLR-3080) We should consider removing shard info from Zk when you explicitly unload a SolrCore.

2012-02-24 Thread Yonik Seeley (Commented) (JIRA)

[ 
https://issues.apache.org/jira/browse/SOLR-3080?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13216244#comment-13216244
 ] 

Yonik Seeley commented on SOLR-3080:


FYI, sometimes the core name and sometimes the zk node name was used to key 
coreStates (leading to duplicate entries on a restart).  I've changed it to 
just key off of core name, along with some other minor improvements like 
creating the map up front so we have a single object to consistently 
synchronize on (prob didn't matter now, but may in the future w/ continued 
development).

 We should consider removing shard info from Zk when you explicitly unload a 
 SolrCore.
 -

 Key: SOLR-3080
 URL: https://issues.apache.org/jira/browse/SOLR-3080
 Project: Solr
  Issue Type: Improvement
  Components: SolrCloud
Reporter: Mark Miller
Assignee: Sami Siren
Priority: Minor
 Fix For: 4.0

 Attachments: SOLR-3080.patch, SOLR-3080.patch, 
 SOLR-3080_smaller_sync.patch




--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: 
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira



-
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org



[jira] [Commented] (SOLR-3080) We should consider removing shard info from Zk when you explicitly unload a SolrCore.

2012-02-14 Thread Mark Miller (Commented) (JIRA)

[ 
https://issues.apache.org/jira/browse/SOLR-3080?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13207909#comment-13207909
 ] 

Mark Miller commented on SOLR-3080:
---

Cool - I'll try and get to this today.

 We should consider removing shard info from Zk when you explicitly unload a 
 SolrCore.
 -

 Key: SOLR-3080
 URL: https://issues.apache.org/jira/browse/SOLR-3080
 Project: Solr
  Issue Type: Improvement
  Components: SolrCloud
Reporter: Mark Miller
Assignee: Mark Miller
Priority: Minor
 Fix For: 4.0

 Attachments: SOLR-3080.patch, SOLR-3080.patch




--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: 
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira



-
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org



[jira] [Commented] (SOLR-3080) We should consider removing shard info from Zk when you explicitly unload a SolrCore.

2012-02-09 Thread Sami Siren (Commented) (JIRA)

[ 
https://issues.apache.org/jira/browse/SOLR-3080?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13204516#comment-13204516
 ] 

Sami Siren commented on SOLR-3080:
--

bq. I doubt it... Probably we should throw an error if you try and create a 
core that already exists?

I checked the wiki at http://wiki.apache.org/solr/CoreAdmin#CREATE and it 
documents this special piece of functionality. I am not sure how that should be 
handled when using solr in cloud mode...

 We should consider removing shard info from Zk when you explicitly unload a 
 SolrCore.
 -

 Key: SOLR-3080
 URL: https://issues.apache.org/jira/browse/SOLR-3080
 Project: Solr
  Issue Type: Improvement
  Components: SolrCloud
Reporter: Mark Miller
Assignee: Mark Miller
Priority: Minor
 Fix For: 4.0

 Attachments: SOLR-3080.patch




--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: 
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira



-
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org



[jira] [Commented] (SOLR-3080) We should consider removing shard info from Zk when you explicitly unload a SolrCore.

2012-02-09 Thread Mark Miller (Commented) (JIRA)

[ 
https://issues.apache.org/jira/browse/SOLR-3080?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13204547#comment-13204547
 ] 

Mark Miller commented on SOLR-3080:
---

bq.  I am not sure how that should be handled when using solr in cloud mode...

Perhaps we should just remove it? I mean, it's a core reload right? Why do we 
need to support that through the create cmd as well?

 We should consider removing shard info from Zk when you explicitly unload a 
 SolrCore.
 -

 Key: SOLR-3080
 URL: https://issues.apache.org/jira/browse/SOLR-3080
 Project: Solr
  Issue Type: Improvement
  Components: SolrCloud
Reporter: Mark Miller
Assignee: Mark Miller
Priority: Minor
 Fix For: 4.0

 Attachments: SOLR-3080.patch




--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: 
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira



-
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org



[jira] [Commented] (SOLR-3080) We should consider removing shard info from Zk when you explicitly unload a SolrCore.

2012-02-09 Thread Sami Siren (Commented) (JIRA)

[ 
https://issues.apache.org/jira/browse/SOLR-3080?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13204549#comment-13204549
 ] 

Sami Siren commented on SOLR-3080:
--

bq. Perhaps we should just remove it?

Sounds good to me.

 We should consider removing shard info from Zk when you explicitly unload a 
 SolrCore.
 -

 Key: SOLR-3080
 URL: https://issues.apache.org/jira/browse/SOLR-3080
 Project: Solr
  Issue Type: Improvement
  Components: SolrCloud
Reporter: Mark Miller
Assignee: Mark Miller
Priority: Minor
 Fix For: 4.0

 Attachments: SOLR-3080.patch




--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: 
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira



-
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org



[jira] [Commented] (SOLR-3080) We should consider removing shard info from Zk when you explicitly unload a SolrCore.

2012-02-09 Thread Yonik Seeley (Commented) (JIRA)

[ 
https://issues.apache.org/jira/browse/SOLR-3080?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13204590#comment-13204590
 ] 

Yonik Seeley commented on SOLR-3080:


bq. I mean, it's a core reload right? Why do we need to support that through 
the create cmd as well?

- it looks like you can specify more stuff on create (you could change the data 
dir, etc).  you could replace that with create-swap-unload but it's 3 calls 
instead of one
- it seems useful to have a create or reload command where you don't have to 
worry if the core was already loaded or not... you just want it to read/reread 
it's config.
 
So I can see how it made sense in the past, but if it causes problems with 
cloud mode then we should do something about that.


 We should consider removing shard info from Zk when you explicitly unload a 
 SolrCore.
 -

 Key: SOLR-3080
 URL: https://issues.apache.org/jira/browse/SOLR-3080
 Project: Solr
  Issue Type: Improvement
  Components: SolrCloud
Reporter: Mark Miller
Assignee: Mark Miller
Priority: Minor
 Fix For: 4.0

 Attachments: SOLR-3080.patch




--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: 
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira



-
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org



[jira] [Commented] (SOLR-3080) We should consider removing shard info from Zk when you explicitly unload a SolrCore.

2012-02-09 Thread Mark Miller (Commented) (JIRA)

[ 
https://issues.apache.org/jira/browse/SOLR-3080?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13204594#comment-13204594
 ] 

Mark Miller commented on SOLR-3080:
---

bq. but if it causes problems with cloud mode then we should do something about 
that.

Well there are other ways to deal with it I imagine - leave it alone and just 
doc not to do that in cloud mode, or if in cloud mode throw an error or 
whatever.

 We should consider removing shard info from Zk when you explicitly unload a 
 SolrCore.
 -

 Key: SOLR-3080
 URL: https://issues.apache.org/jira/browse/SOLR-3080
 Project: Solr
  Issue Type: Improvement
  Components: SolrCloud
Reporter: Mark Miller
Assignee: Mark Miller
Priority: Minor
 Fix For: 4.0

 Attachments: SOLR-3080.patch




--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: 
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira



-
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org



[jira] [Commented] (SOLR-3080) We should consider removing shard info from Zk when you explicitly unload a SolrCore.

2012-02-08 Thread Mark Miller (Commented) (JIRA)

[ 
https://issues.apache.org/jira/browse/SOLR-3080?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13204114#comment-13204114
 ] 

Mark Miller commented on SOLR-3080:
---

{quote}I noticed that CoreAdminHandler allows creating multiple cores with 
same name and if you do that you can end up having one core in multiple places 
unless you remove it in between. Is this behavior of CoreAdminHandler 
intentional?{quote}

I doubt it... Probably we should throw an error if you try and create a core 
that already exists?

 We should consider removing shard info from Zk when you explicitly unload a 
 SolrCore.
 -

 Key: SOLR-3080
 URL: https://issues.apache.org/jira/browse/SOLR-3080
 Project: Solr
  Issue Type: Improvement
  Components: SolrCloud
Reporter: Mark Miller
Assignee: Mark Miller
Priority: Minor
 Fix For: 4.0

 Attachments: SOLR-3080.patch




--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: 
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira



-
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org



[jira] [Commented] (SOLR-3080) We should consider removing shard info from Zk when you explicitly unload a SolrCore.

2012-02-03 Thread Mark Miller (Commented) (JIRA)

[ 
https://issues.apache.org/jira/browse/SOLR-3080?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13199754#comment-13199754
 ] 

Mark Miller commented on SOLR-3080:
---

bq. Also I was not sure if the slice should be deleted when there are no more 
shards with that id (currently it is not deleted).

Yeah, I think that is the right move for now - don't delete it.

 We should consider removing shard info from Zk when you explicitly unload a 
 SolrCore.
 -

 Key: SOLR-3080
 URL: https://issues.apache.org/jira/browse/SOLR-3080
 Project: Solr
  Issue Type: Improvement
  Components: SolrCloud
Reporter: Mark Miller
Assignee: Mark Miller
Priority: Minor
 Fix For: 4.0

 Attachments: SOLR-3080.patch




--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: 
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira



-
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org



[jira] [Commented] (SOLR-3080) We should consider removing shard info from Zk when you explicitly unload a SolrCore.

2012-02-03 Thread Mark Miller (Commented) (JIRA)

[ 
https://issues.apache.org/jira/browse/SOLR-3080?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13199755#comment-13199755
 ] 

Mark Miller commented on SOLR-3080:
---

bq. Also I was not sure if the slice should be deleted when there are no more 
shards with that id (currently it is not deleted).

Yeah, I think that is the right move for now - don't delete it.

 We should consider removing shard info from Zk when you explicitly unload a 
 SolrCore.
 -

 Key: SOLR-3080
 URL: https://issues.apache.org/jira/browse/SOLR-3080
 Project: Solr
  Issue Type: Improvement
  Components: SolrCloud
Reporter: Mark Miller
Assignee: Mark Miller
Priority: Minor
 Fix For: 4.0

 Attachments: SOLR-3080.patch




--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: 
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira



-
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org