You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@tinkerpop.apache.org by GitBox <gi...@apache.org> on 2020/10/08 22:02:04 UTC

[GitHub] [tinkerpop] divijvaidya opened a new pull request #1343: Tinkerpop 2440 Simplify driver by delegating keepAlive logic to Netty

divijvaidya opened a new pull request #1343:
URL: https://github.com/apache/tinkerpop/pull/1343


   https://issues.apache.org/jira/browse/TINKERPOP-2440
   
   Java driver can be simplified by delegating the keepAlive logic to Netty's IdleStateHandler instead of maintaining and writing our own custom code. 
   
   Note that this implementation is backward compatible for 3.4.x series as the existing keepAlive logic should still continue to work for Channelizers other than `WebSocketChannelizer` (notice the if statement in the code that checks for channelizer class before scheduling keepAlive). For `WebSocketChannelizer`, the number of Ping frames sent to the server is still the same except that it is now handled by Netty.
   
   ### Testing
   1. Added a new unit test
   2. gremlin-server `mvn clean install -DskipIntegrationTests=false -DincludeNeo4j`
   3. gremlin-driver `mvn clean install -DskipIntegrationTests=false -DincludeNeo4j`
   
   


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [tinkerpop] spmallette commented on pull request #1343: TINKERPOP-2440 Simplify driver by delegating keepAlive logic to Netty

Posted by GitBox <gi...@apache.org>.
spmallette commented on pull request #1343:
URL: https://github.com/apache/tinkerpop/pull/1343#issuecomment-712756886


   hmm - true. perhaps some aspect of keep-alive was wrong from the beginning. The `Connection` probably shouldn't have needed to know anything about keep-alive infrastructure. If you agree, I think I'd propose that we deprecate that functionality and remove it in 3.5.0 (all of this can be in a different JIRA and we can move forward with merging this as it is). Does that make sense?
   
   Either way, VOTE +1


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [tinkerpop] spmallette commented on pull request #1343: TINKERPOP-2440 Simplify driver by delegating keepAlive logic to Netty

Posted by GitBox <gi...@apache.org>.
spmallette commented on pull request #1343:
URL: https://github.com/apache/tinkerpop/pull/1343#issuecomment-707653519


   Good change - Glad the `shouldKeepAliveForWebSockets()` test is back. I suppose there is nothing now testing `scheduleKeepAlive()` logic - which remains fairly complex. I haven't thought about it, but is there any way to continue to cover that logic with tests (until we, i assume, one day remove the custom keep-alive all together)? 


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [tinkerpop] divijvaidya commented on pull request #1343: TINKERPOP-2440 Simplify driver by delegating keepAlive logic to Netty

Posted by GitBox <gi...@apache.org>.
divijvaidya commented on pull request #1343:
URL: https://github.com/apache/tinkerpop/pull/1343#issuecomment-712500482


   > Good change - Glad the `shouldKeepAliveForWebSockets()` test is back. I suppose there is nothing now testing `scheduleKeepAlive()` logic - which remains fairly complex. I haven't thought about it, but is there any way to continue to cover that logic with tests (until we, i assume, one day remove the custom keep-alive all together)?
   
   Let's keep this out of the scope of this review? Asking because we don't even have a channelizer in the code base which would use the old code path, hence, testing becomes difficult.


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [tinkerpop] spmallette commented on pull request #1343: TINKERPOP-2440 Simplify driver by delegating keepAlive logic to Netty

Posted by GitBox <gi...@apache.org>.
spmallette commented on pull request #1343:
URL: https://github.com/apache/tinkerpop/pull/1343#issuecomment-713481848


   I made some minor tweaks to deprecation messaging: 35871ebdbbfd9799f230d37013da511a9487672d and created TINKERPOP-2455 for removal of the functionality in 3.5.0


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [tinkerpop] divijvaidya commented on pull request #1343: TINKERPOP-2440 Simplify driver by delegating keepAlive logic to Netty

Posted by GitBox <gi...@apache.org>.
divijvaidya commented on pull request #1343:
URL: https://github.com/apache/tinkerpop/pull/1343#issuecomment-713014082


   > hmm - true. perhaps some aspect of keep-alive was wrong from the beginning. The `Connection` probably shouldn't have needed to know anything about keep-alive infrastructure. If you agree, I think I'd propose that we deprecate that functionality and remove it in 3.5.0 (all of this can be in a different JIRA and we can move forward with merging this as it is). Does that make sense?
   > 
   > Either way, VOTE +1
   
   Yes I agree that this should be deprecated in 3.5 I will remove it from master as part of the merge and do a CTR update to documentation.


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [tinkerpop] spmallette edited a comment on pull request #1343: TINKERPOP-2440 Simplify driver by delegating keepAlive logic to Netty

Posted by GitBox <gi...@apache.org>.
spmallette edited a comment on pull request #1343:
URL: https://github.com/apache/tinkerpop/pull/1343#issuecomment-713481848


   I made some minor tweaks to deprecation messaging: 35871ebdbbfd9799f230d37013da511a9487672d and created  https://issues.apache.org/jira/browse/TINKERPOP-2455 for removal of the functionality in 3.5.0


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



[GitHub] [tinkerpop] divijvaidya merged pull request #1343: TINKERPOP-2440 Simplify driver by delegating keepAlive logic to Netty

Posted by GitBox <gi...@apache.org>.
divijvaidya merged pull request #1343:
URL: https://github.com/apache/tinkerpop/pull/1343


   


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org