Maximum executor retries limit in spark worker

[mirjan....@gmail.com]

Hi everyone!
How to limit maximum executor retries limit in spark worker?

I tried:

{
 "engineId": "rb",
"engineFactory": "com.actionml.engines.ur.UREngine",
"sparkConf": {
"master": "spark://spark-master:7077",
"spark.serializer": "org.apache.spark.serializer.KryoSerializer",
"spark.kryo.registrator": "org.apache.mahout.sparkbindings.io.MahoutKryoRegistrator",
"spark.kryo.referenceTracking": "false",
"spark.kryoserializer.buffer": "300m",
"spark.executor.memory": "2g",
"spark.driver.memory": "2g",
"spark.es.index.auto.create": "true",
"spark.es.nodes": "elasticsearch",
"spark.es.nodes.wan.only": "true",
"spark.dynamicAllocation.maxExecutors": "3",
"spark.worker.cleanup.enabled": "true"
},
"algorithm":{
"indicators": [
{
"name": "buy"
},
{
"name": "add-to-cart"
},
{
"name": "view"
},
{
"name": "like"
}
]
}
}

But it does not work. 

[mirjan....@gmail.com]

[Pat Ferrel]

Harness expects Spark 2.3.x You seem to be running a newer version 2.4.5, which may cause problems — not sure.

What does the main Spark GUI screen show? How many cores are running Spark Workers?

From: mirjan....@gmail.com <mirjan....@gmail.com>
Date: November 8, 2020 at 11:59:01 PM
To: actionml-user <action...@googlegroups.com>
Subject:  Re: Maximum executor retries limit in spark worker

On Monday, November 9, 2020 at 11:55:38 AM UTC+4 mirjan....@gmail.com wrote:

--
You received this message because you are subscribed to the Google Groups "actionml-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email to actionml-use...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/actionml-user/82a5953b-1b06-4326-a489-104dc38104can%40googlegroups.com.