juliuszsompolski commented on code in PR #41829:
URL: https://github.com/apache/spark/pull/41829#discussion_r1254255715
##########
connector/connect/client/jvm/src/main/scala/org/apache/spark/sql/connect/client/ArtifactManager.scala:
##########
@@ -51,13 +51,14 @@ import org.apache.spark.util.{SparkFileUtils,
SparkThreadUtils}
class ArtifactManager(
userContext: proto.UserContext,
sessionId: String,
- channel: ManagedChannel) {
+ channel: ManagedChannel,
+ retryPolicy: GrpcRetryHandler.RetryPolicy) {
// Using the midpoint recommendation of 32KiB for chunk size as specified in
// https://github.com/grpc/grpc.github.io/issues/371.
private val CHUNK_SIZE: Int = 32 * 1024
- private[this] val stub = proto.SparkConnectServiceGrpc.newStub(channel)
- private[this] val bstub =
proto.SparkConnectServiceGrpc.newBlockingStub(channel)
+ private[this] val stub = new CustomSparkConnectStub(channel, retryPolicy)
+ private[this] val bstub = new CustomSparkConnectBlockingStub(channel,
retryPolicy)
Review Comment:
I see. I was thinking of having the client class as a single narrow waist
for all rpc communication, without analyzing too much whats in ArtifactManager,
but looking how much is there, I think you're right.
Let's create the stubs in the client and pass these to ArtifactManager like
you suggest.
--
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.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
