You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@tinkerpop.apache.org by "ASF GitHub Bot (JIRA)" <ji...@apache.org> on 2016/03/02 00:22:18 UTC
[jira] [Commented] (TINKERPOP-956) Connection errors tend to force
a complete close of the channel
[ https://issues.apache.org/jira/browse/TINKERPOP-956?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15174609#comment-15174609 ]
ASF GitHub Bot commented on TINKERPOP-956:
------------------------------------------
GitHub user spmallette opened a pull request:
https://github.com/apache/incubator-tinkerpop/pull/248
TINKERPOP-956 Connection errors tend to force a complete close of the channel
https://issues.apache.org/jira/browse/TINKERPOP-956
Implemented a method of refreshing a connection when notification was received that it was closed on the server-side. As a result the driver should do a better job of recovering in the face of a server-side closing of a channel.
Had a good build with `mvn clean install && mvn verify -DskipIntegrationTests=false -DincludeNeo4j`
VOTE +1
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/apache/incubator-tinkerpop TINKERPOP-956
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/incubator-tinkerpop/pull/248.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #248
----
commit fba51aa14afca428a1819685f8313be1b0ec3526
Author: Stephen Mallette <sp...@genoprime.com>
Date: 2016-03-01T20:44:44Z
General fix to allow server-side fails to be reconnected.
If a server is responding to the driver in an expected fashion, then the server is still valid, but it may have killed the channel. If it killed the channel and the driver tries to reconnect on that channel then future requests on that connection would immediately fail and the host would then get marked as dead (when it really wasn't). Made some basic adjustments here to allow a Connection facing this situation to replace itself in the pool with a fresh connection which would open a new channel on the server and thus allow the next request using it to actually work.
commit 4abb58bf5df63b71d2ccab893f7f9f354e2391a1
Author: Stephen Mallette <sp...@genoprime.com>
Date: 2016-03-01T21:52:27Z
Minor refactoring to extract Connection shutdown function.
commit 9e1b5b12da38f0c13573cbea25a94f18a4cab104
Author: Stephen Mallette <sp...@genoprime.com>
Date: 2016-03-01T22:33:49Z
Ignored the shouldHandleRequestSentThatNeverReturns() test for driver.
This test never wants to behave not matter is done to it to firm it up. It seems to rely a little too heavily on timing to be consistent.
commit 0ff4e8285a939e75c66636057c9cc287cc1cba64
Author: Stephen Mallette <sp...@genoprime.com>
Date: 2016-03-01T22:56:42Z
Update changelog.
----
> Connection errors tend to force a complete close of the channel
> ---------------------------------------------------------------
>
> Key: TINKERPOP-956
> URL: https://issues.apache.org/jira/browse/TINKERPOP-956
> Project: TinkerPop
> Issue Type: Improvement
> Components: driver, server
> Affects Versions: 3.0.2-incubating
> Reporter: stephen mallette
> Assignee: stephen mallette
> Priority: Minor
> Fix For: 3.1.2-incubating
>
>
> Would be nice if the channel could stay open if the error is recoverable.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)