[ https://issues.apache.org/jira/browse/SPARK-43366?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17726731#comment-17726731 ]
Sean R. Owen commented on SPARK-43366: -------------------------------------- Ack yeah, nevermind. Reading too fast without coffee. That I don't know, except to say that's not going to be controlled by the _driver_ IP. Block manager would be tied to _executors_. Is that IP an executor? > Spark Driver Bind Address is off-by-one > --------------------------------------- > > Key: SPARK-43366 > URL: https://issues.apache.org/jira/browse/SPARK-43366 > Project: Spark > Issue Type: Bug > Components: Block Manager > Affects Versions: 3.3.3 > Reporter: Derek Brown > Priority: Major > > I have the following environment variable set in my driver pod configuration: > {code:java} > SPARK_DRIVER_BIND_ADDRESS=10.244.0.53{code} > However, I see an off-by-one IP address being referred to in the Spark logs: > {code:java} > 23/05/04 02:37:03 INFO > KubernetesClusterSchedulerBackend$KubernetesDriverEndpoint: Registered > executor NettyRpcEndpointRef(spark-client://Executor) (10.244.0.54:53140) > with ID 1, ResourceProfileId 0 > 23/05/04 02:37:03 INFO BlockManagerMasterEndpoint: Registering block manager > 10.244.0.54:32805 with 413.9 MiB RAM, BlockManagerId(1, 10.244.0.54, 32805, > None){code} > I am not sure why this might be the case. -- This message was sent by Atlassian Jira (v8.20.10#820010) --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org