[
https://issues.apache.org/jira/browse/HIVE-18958?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16454284#comment-16454284
]
Sahil Takiar commented on HIVE-18958:
-------------------------------------
No, spark.yarn.driver.memoryOverhead and spark.yarn.driver.memoryOverhead are
Spark configs that were renamed in Spark 2.x, Hive is just a user of those
configs, not the owner. So the changes you made won't cause any
incompatibilities.
A few minor comments, overall approach looks good:
* Can you make all the {{\-\-}} parameters defined in the {{RemoteDriver}}
constructor static strings - e.g. {{\-\-remote-host}} should be a {{public
static final String REMOTE_HOST_CONF = "--remote-host"}}; both the
{{RemoteDriver}} and {{SparkClientImpl}} should use this variable; this is more
of a style issue, constants should be defined as static variables
* Can you rename the folder {{data/conf/spark/yarn-client}} to
{{data/conf/spark/yarn-cluster}}, we are actually running Spark in {{cluster}}
mode not {{client}} mode (as defined by the config {{spark.submit.deployMode}}
in {{data/conf/spark/yarn-client/hive-site.xml}})
> Fix Spark config warnings
> -------------------------
>
> Key: HIVE-18958
> URL: https://issues.apache.org/jira/browse/HIVE-18958
> Project: Hive
> Issue Type: Sub-task
> Components: Spark
> Reporter: Sahil Takiar
> Assignee: Bharathkrishna Guruvayoor Murali
> Priority: Major
> Attachments: HIVE-18958.01.patch, HIVE-18958.02.patch
>
>
> Getting a few configuration warnings in the logs that we should fix:
> {code}
> 2018-03-14T10:06:19,164 WARN [d5ade9e4-9354-40f1-8f74-631f373709b3 main]
> spark.SparkConf: The configuration key 'spark.yarn.driver.memoryOverhead' has
> been deprecated as of Spark 2.3 and may be removed in the future. Please use
> the new key 'spark.driver.memoryOverhead' instead.
> 2018-03-14T10:06:19,165 WARN [d5ade9e4-9354-40f1-8f74-631f373709b3 main]
> spark.SparkConf: The configuration key spark.akka.logLifecycleEvents is not
> supported any more because Spark doesn't use Akka since 2.0
> 2018-03-14T10:06:19,165 WARN [d5ade9e4-9354-40f1-8f74-631f373709b3 main]
> spark.SparkConf: The configuration key 'spark.yarn.executor.memoryOverhead'
> has been deprecated as of Spark 2.3 and may be removed in the future. Please
> use the new key 'spark.executor.memoryOverhead' instead.
> 2018-03-14T10:06:20,351 INFO
> [RemoteDriver-stderr-redir-d5ade9e4-9354-40f1-8f74-631f373709b3 main]
> client.SparkClientImpl: Warning: Ignoring non-spark config property:
> hive.spark.client.server.connect.timeout=90000
> 2018-03-14T10:06:20,351 INFO
> [RemoteDriver-stderr-redir-d5ade9e4-9354-40f1-8f74-631f373709b3 main]
> client.SparkClientImpl: Warning: Ignoring non-spark config property:
> hive.spark.client.rpc.threads=8
> 2018-03-14T10:06:20,351 INFO
> [RemoteDriver-stderr-redir-d5ade9e4-9354-40f1-8f74-631f373709b3 main]
> client.SparkClientImpl: Warning: Ignoring non-spark config property:
> hive.spark.client.connect.timeout=30000
> 2018-03-14T10:06:20,351 INFO
> [RemoteDriver-stderr-redir-d5ade9e4-9354-40f1-8f74-631f373709b3 main]
> client.SparkClientImpl: Warning: Ignoring non-spark config property:
> hive.spark.client.secret.bits=256
> 2018-03-14T10:06:20,351 INFO
> [RemoteDriver-stderr-redir-d5ade9e4-9354-40f1-8f74-631f373709b3 main]
> client.SparkClientImpl: Warning: Ignoring non-spark config property:
> hive.spark.client.rpc.max.size=52428800
> {code}
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
