Dan Burkert has posted comments on this change.

Change subject: Statically cache kudu connections This avoids connection leak 
issues when running large spark jobs
......................................................................


Patch Set 1:

(2 comments)

http://gerrit.cloudera.org:8080/#/c/3124/1/java/kudu-spark/src/main/scala/org/kududb/spark/kudu/KuduContext.scala
File java/kudu-spark/src/main/scala/org/kududb/spark/kudu/KuduContext.scala:

Line 50:     ShutdownHookManager.get().addShutdownHook(new Runnable {
I think the two shutdown hook adds here and for the async version should be 
moved to inside the synchronized blocks below, otherwise it could register 
multiple shutdown hooks per client.


Line 84:   val syncCache = new mutable.HashMap[String, KuduClient]()
could these fields be private?  It may not make a difference since the object 
is at private scope.


-- 
To view, visit http://gerrit.cloudera.org:8080/3124
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-MessageType: comment
Gerrit-Change-Id: Ic4916ed47208020e5c91c4280c73f9dbd7e3d54f
Gerrit-PatchSet: 1
Gerrit-Project: kudu
Gerrit-Branch: master
Gerrit-Owner: Brent Gardner <[email protected]>
Gerrit-Reviewer: Dan Burkert <[email protected]>
Gerrit-Reviewer: Kudu Jenkins
Gerrit-HasComments: Yes

Reply via email to