Jean-Daniel Cryans has posted comments on this change. Change subject: spark: add support for fault tolerant scanner ......................................................................
Patch Set 3: (2 comments) http://gerrit.cloudera.org:8080/#/c/6782/1/java/kudu-spark/src/main/scala/org/apache/kudu/spark/kudu/DefaultSource.scala File java/kudu-spark/src/main/scala/org/apache/kudu/spark/kudu/DefaultSource.scala: PS1, Line 100: getScannerValue I think doing something like "Try(parameters.getOrElse(FAULT_TOLERANT_SCANNER, "false").toBoolean)).getOrElse(false)" would negate the need for a new method. http://gerrit.cloudera.org:8080/#/c/6782/1/java/kudu-spark/src/main/scala/org/apache/kudu/spark/kudu/KuduContext.scala File java/kudu-spark/src/main/scala/org/apache/kudu/spark/kudu/KuduContext.scala: Line 43: class KuduContext(kuduMaster: String, I think setting this on the context is too restrictive, it means you'd have to keep two sets of connections if you want to do a mix of fault tolerant and non fault tolerant scanners. I think a better choice would be to do it in the KuduRDD ctor. Also, wouldn't this current change be backward-incompatible? -- To view, visit http://gerrit.cloudera.org:8080/6782 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: I3f3192025ca5d74197600480fd3d040d70b4bbc2 Gerrit-PatchSet: 3 Gerrit-Project: kudu Gerrit-Branch: master Gerrit-Owner: Hao Hao <hao....@cloudera.com> Gerrit-Reviewer: Dan Burkert <danburk...@apache.org> Gerrit-Reviewer: Jean-Daniel Cryans <jdcry...@apache.org> Gerrit-Reviewer: Kudu Jenkins Gerrit-HasComments: Yes
