I want to read data from one collection on a sharding cluster, after some transformation, write to another collection on a standalone server. Spark config is like below:

```

String iputUri = "mongodb://usr:pwd@ip1:port1,ip2:port2,ip3:port3,ip4:port4,ip5:port5,ip6:port6/db.col?maxPoolSize=40&authSource=admin&readPreference=secondary";

String outputUri = "mongodb://usr2:pw2@ip7:port7/db2.col2?maxPoolSize=40&authSource=admin&readPreference=secondary";

SparkSession sparkR =BUILDER.appName("myapp")
.config("spark.mongodb.input.uri", inputUri)

.config("spark.mongodb.output.uri", ouputUri)
.config("spark.mongodb.input.partitionerOptions.partitionKey", "id")
.config("spark.mongodb.input.partitioner", "MongoSamplePartitioner")
.getOrCreate();

```

Exception occurred when executing task: Command failed with error 13: 'not authorized on db2 to execute command { update: "col2", ordered: true, writeConcern:

{ w: 1 }

, $db: "db2", $clusterTime: { clusterTime: Timestamp(1574246888, 5), signature:

{ hash: BinData(0, DB1EDCE1A9DA107BFE21A700396763ADBF09D528), keyId: 6710554364668280834 }

}, lsid: { id: UUID("952698d2-b4ae-40b8-beff-354de32a9f8c") } }' on server ip2:port2. 

Then I changed codes to put output uri in WriteConfig:

```

Map<String, String> writeOverrides = new HashMap<>();
writeOverrides.put("writeConcern.w", "1");
writeOverrides.put("replaceDocument", "true");
writeOverrides.put("uri", outputUri);
WriteConfig writeConfig = WriteConfig.create(JSC).withOptions(writeOverrides);

```

Still failed with com.mongodb.MongoInterruptedException: Interrupted acquiring a permit to retrieve an item from the pool.

And the version is:

```

compile 'org.mongodb.spark:mongo-spark-connector_2.11:2.1.5'

```