I am using the azure-cosmosdb-spark connector to write data into cosmosdb from cloudera spark cluster. The connectivity works well and the documents gets created in the cosmos db.
But when I am running the Spark application in client mode, the spark application does not stop even though the processing has finished and the spark session has ended.
The spark driver session in the client mode is kept active by the spark.CosmosDBConnectionCache: throughput-refresh-timer.
The info logs related to spark.CosmosDBConnectionCache: throughput-refresh-timer timer refresh keeps on coming and Spark Application does not finish itself.
I tried using System.exit(0) and it works, the application completes sucessfully. But without this, it does not exits.
Below are the logs which keep on printing:
20/09/01 19:18:45 INFO spark.CosmosDBConnectionCache: throughput-refresh-timer: ClientConfiguration#-631429096 has been reset - new ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']), previously ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']) 20/09/01 19:33:45 INFO spark.CosmosDBConnectionCache: throughput-refresh-timer: ClientConfiguration#-631429096 has been reset - new ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']), previously ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']) 20/09/01 19:48:45 INFO spark.CosmosDBConnectionCache: throughput-refresh-timer: ClientConfiguration#-631429096 has been reset - new ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']), previously ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']) 20/09/01 20:03:45 INFO spark.CosmosDBConnectionCache: throughput-refresh-timer: ClientConfiguration#-631429096 has been reset - new ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']), previously ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']) 20/09/01 20:18:45 INFO spark.CosmosDBConnectionCache: throughput-refresh-timer: ClientConfiguration#-631429096 has been reset - new ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']), previously ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']) 20/09/01 20:33:45 INFO spark.CosmosDBConnectionCache: throughput-refresh-timer: ClientConfiguration#-631429096 has been reset - new ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']), previously ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']) 20/09/01 20:48:45 INFO spark.CosmosDBConnectionCache: throughput-refresh-timer: ClientConfiguration#-631429096 has been reset - new ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']), previously ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s'])
I am using the
azure-cosmosdb-spark connector
to write data into cosmosdb from cloudera spark cluster. The connectivity works well and the documents gets created in the cosmos db.But when I am running the Spark application in
client
mode, the spark application does not stop even though the processing has finished and the spark session has ended.The spark driver session in the client mode is kept active by the
spark.CosmosDBConnectionCache: throughput-refresh-timer
. The info logs related tospark.CosmosDBConnectionCache: throughput-refresh-timer
timer refresh keeps on coming and Spark Application does not finish itself.I tried using
System.exit(0)
and it works, the application completes sucessfully. But without this, it does not exits. Below are the logs which keep on printing:20/09/01 19:18:45 INFO spark.CosmosDBConnectionCache: throughput-refresh-timer: ClientConfiguration#-631429096 has been reset - new ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']), previously ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']) 20/09/01 19:33:45 INFO spark.CosmosDBConnectionCache: throughput-refresh-timer: ClientConfiguration#-631429096 has been reset - new ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']), previously ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']) 20/09/01 19:48:45 INFO spark.CosmosDBConnectionCache: throughput-refresh-timer: ClientConfiguration#-631429096 has been reset - new ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']), previously ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']) 20/09/01 20:03:45 INFO spark.CosmosDBConnectionCache: throughput-refresh-timer: ClientConfiguration#-631429096 has been reset - new ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']), previously ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']) 20/09/01 20:18:45 INFO spark.CosmosDBConnectionCache: throughput-refresh-timer: ClientConfiguration#-631429096 has been reset - new ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']), previously ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']) 20/09/01 20:33:45 INFO spark.CosmosDBConnectionCache: throughput-refresh-timer: ClientConfiguration#-631429096 has been reset - new ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']), previously ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']) 20/09/01 20:48:45 INFO spark.CosmosDBConnectionCache: throughput-refresh-timer: ClientConfiguration#-631429096 has been reset - new ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s']), previously ClientCacheEntry#1640518163(DocumentClient#702954435, ContainerMetadata#-561333454['n***************s'])