You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@cassandra.apache.org by "Chris Goffinet (JIRA)" <ji...@apache.org> on 2009/12/19 23:26:18 UTC
[jira] Created: (CASSANDRA-648) findSuitableEndPoint not returning
closest endPoint
findSuitableEndPoint not returning closest endPoint
---------------------------------------------------
Key: CASSANDRA-648
URL: https://issues.apache.org/jira/browse/CASSANDRA-648
Project: Cassandra
Issue Type: Bug
Affects Versions: 0.5
Reporter: Chris Goffinet
Assignee: Chris Goffinet
Fix For: 0.5
After we switched everything over to Rackaware, I was noticing queries still taking 80ms+. We run in multiple datacenter's and I had a suspicion that it was just connecting to a node on east coast. It looks like findSuitableEndPoint sorts endpoints, but does not use the returned sorted endpoints when needing to return the final endpoint.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (CASSANDRA-648) findSuitableEndPoint not
returning closest endPoint
Posted by "Chris Goffinet (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-648?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12793279#action_12793279 ]
Chris Goffinet commented on CASSANDRA-648:
------------------------------------------
+1. Can you add the javadoc that Stu asked about for sortByProximity?
> findSuitableEndPoint not returning closest endPoint
> ---------------------------------------------------
>
> Key: CASSANDRA-648
> URL: https://issues.apache.org/jira/browse/CASSANDRA-648
> Project: Cassandra
> Issue Type: Bug
> Affects Versions: 0.5
> Reporter: Chris Goffinet
> Assignee: Jonathan Ellis
> Fix For: 0.5
>
> Attachments: 0001-Returned-the-sorted-endpoints-properly-for-CASSANDRA.patch, 648-2.patch, 648.patch
>
>
> After we switched everything over to Rackaware, I was noticing queries still taking 80ms+. We run in multiple datacenter's and I had a suspicion that it was just connecting to a node on east coast. It looks like findSuitableEndPoint sorts endpoints, but does not use the returned sorted endpoints when needing to return the final endpoint.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (CASSANDRA-648) findSuitableEndPoint not returning
closest endPoint
Posted by "Jonathan Ellis (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-648?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Jonathan Ellis updated CASSANDRA-648:
-------------------------------------
Attachment: 648.patch
For efficiency, I'd prefer to make sortByProximity work in-place as its name suggests.
But we need the non-in-place for the Multimaps we have.
So here is a patch w/ two versions, the one leveraging the other of course.
> findSuitableEndPoint not returning closest endPoint
> ---------------------------------------------------
>
> Key: CASSANDRA-648
> URL: https://issues.apache.org/jira/browse/CASSANDRA-648
> Project: Cassandra
> Issue Type: Bug
> Affects Versions: 0.5
> Reporter: Chris Goffinet
> Assignee: Chris Goffinet
> Fix For: 0.5
>
> Attachments: 0001-Returned-the-sorted-endpoints-properly-for-CASSANDRA.patch, 648.patch
>
>
> After we switched everything over to Rackaware, I was noticing queries still taking 80ms+. We run in multiple datacenter's and I had a suspicion that it was just connecting to a node on east coast. It looks like findSuitableEndPoint sorts endpoints, but does not use the returned sorted endpoints when needing to return the final endpoint.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (CASSANDRA-648) findSuitableEndPoint not returning
closest endPoint
Posted by "Chris Goffinet (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-648?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Chris Goffinet updated CASSANDRA-648:
-------------------------------------
Attachment: 0001-Returned-the-sorted-endpoints-properly-for-CASSANDRA.patch
> findSuitableEndPoint not returning closest endPoint
> ---------------------------------------------------
>
> Key: CASSANDRA-648
> URL: https://issues.apache.org/jira/browse/CASSANDRA-648
> Project: Cassandra
> Issue Type: Bug
> Affects Versions: 0.5
> Reporter: Chris Goffinet
> Assignee: Chris Goffinet
> Fix For: 0.5
>
> Attachments: 0001-Returned-the-sorted-endpoints-properly-for-CASSANDRA.patch
>
>
> After we switched everything over to Rackaware, I was noticing queries still taking 80ms+. We run in multiple datacenter's and I had a suspicion that it was just connecting to a node on east coast. It looks like findSuitableEndPoint sorts endpoints, but does not use the returned sorted endpoints when needing to return the final endpoint.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (CASSANDRA-648) findSuitableEndPoint not
returning closest endPoint
Posted by "Stu Hood (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-648?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12792919#action_12792919 ]
Stu Hood commented on CASSANDRA-648:
------------------------------------
+1... but before you commit, can you add a javadoc for that method, so we can end the cycle?
Thanks!
> findSuitableEndPoint not returning closest endPoint
> ---------------------------------------------------
>
> Key: CASSANDRA-648
> URL: https://issues.apache.org/jira/browse/CASSANDRA-648
> Project: Cassandra
> Issue Type: Bug
> Affects Versions: 0.5
> Reporter: Chris Goffinet
> Assignee: Chris Goffinet
> Fix For: 0.5
>
> Attachments: 0001-Returned-the-sorted-endpoints-properly-for-CASSANDRA.patch
>
>
> After we switched everything over to Rackaware, I was noticing queries still taking 80ms+. We run in multiple datacenter's and I had a suspicion that it was just connecting to a node on east coast. It looks like findSuitableEndPoint sorts endpoints, but does not use the returned sorted endpoints when needing to return the final endpoint.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (CASSANDRA-648) findSuitableEndPoint not returning
closest endPoint
Posted by "Jonathan Ellis (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-648?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Jonathan Ellis updated CASSANDRA-648:
-------------------------------------
Attachment: 648-2.patch
move isInSameRack and isInSameDC into AbstractEPS, making the basic IEPS interface sorting by proximity and the rest an implementation detail that only the corresponding kind of ReplicationStrategy should care about
> findSuitableEndPoint not returning closest endPoint
> ---------------------------------------------------
>
> Key: CASSANDRA-648
> URL: https://issues.apache.org/jira/browse/CASSANDRA-648
> Project: Cassandra
> Issue Type: Bug
> Affects Versions: 0.5
> Reporter: Chris Goffinet
> Assignee: Chris Goffinet
> Fix For: 0.5
>
> Attachments: 0001-Returned-the-sorted-endpoints-properly-for-CASSANDRA.patch, 648-2.patch, 648.patch
>
>
> After we switched everything over to Rackaware, I was noticing queries still taking 80ms+. We run in multiple datacenter's and I had a suspicion that it was just connecting to a node on east coast. It looks like findSuitableEndPoint sorts endpoints, but does not use the returned sorted endpoints when needing to return the final endpoint.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.