QueryDB

To provide high availability or load balancing for HiveServer2, Hive provides a function called dynamic service discovery where multiple HiveServer2 instances can register themselves with Zookeeper. Instead of connecting to a specific HiveServer2 directly, clients connect to Zookeeper which returns a randomly selected registered HiveServer2 instance. For example -  Below command connects to Hive Server on MachineA beeline -u "jdbc:hive2://machineA: 10000" Below command connects to Zookeeper Node: to determine one of the available Hive Server's to make a connection beeline -u "jdbc:hive2://machineA:2181,machineB:2181,machineC:2181/;serviceDiscoveryMode=zooKeeper;zooKeeperNamespace=hiveserver2-mob-batch?tez.queue.name=myyarnqueue" We can Create ZNode with Zookeeper as follows -  Open Zookeeper command line interface zookeeper-client Connect to Zookeeper Server connect  machineA:2181,machineB:2181,machineC:2181 Create ZNode create / hiveserver2-mob-batch Manually,

  We were running an application which was leading to below error -  Job aborted due to stage failure: Task 137 in stage 5.0 failed 4 times, most recent failure: Lost task 137.3 in stage 5.0 (TID 2090, ncABC.hadoop.com, executor 1): org.apache.spark.SparkException: Kryo serialization failed: Buffer overflow. Available: 0, required: 59606960. To avoid this, increase spark.kryoserializer.buffer.max value. at org.apache.spark.serializer.KryoSerializerInstance.serialize(KryoSerializer.scala:330) at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:456) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) at java.lang.Thread.run(Thread.java:748) Caused by: com.esotericsoftware.kryo.KryoException: Buffer overflow. Available: 0, required: 59606960 at com.esotericsoftware.kryo.io.Output.require(Output.java:167) at com.esotericsoftware.kryo.io.Output.writeBytes(

  Spark MongoDB connector does not fetch all the fields present in stored BSON document in a collection.  This is because Mongo collection can have documents with different schema. Typically, all documents in a collection are of similar or related purpose. A document is a set of key-value pairs. Documents have dynamic schema. Dynamic schema means that documents in the same collection do not need to have the same set of fields or structure, and common fields in a collection's documents may hold different types of data. So, when we read MongoDB Collection using Spark connector. It infers schema as per first row it may read, which might not consist of fields which are present in subsequent tuples/ rows. Suppose We have a MongoDB collection - default.fruits, and it has following documents -  { "_id" : 1, "type" : "apple"} { "_id" : 2, "type" : "orange", "qty" : 10 } { "_id" : 3, "type" : "ban

  Easiest available way is using Openssl Encrypt  Say your actual password is "Password12345". You can encrypt using below command - echo "Password12345" | openssl enc -aes-256-cbc -md sha512 -a -pbkdf2 -iter 100000  -salt -pass pass:Secret@1234# > secret.txt Resultant output in " secret.txt" U2FsdGVkX1/2NoQ6i1uZKK4yk+5gm5cA13EJ2TiPbcw= Save this in a file and use it with your applications. Decrypt  Then you can read file with encrypted password, decrypt it using below command and pass it to henceforth operations/ commands. cat "secret.txt" | openssl enc -aes-256-cbc -md sha512 -a -d -pbkdf2 -iter 100000 -salt -pass pass:Secret@123# Resultant output  Password12345

  We had a Spark Job which was taking over 3 hours to complete.  First, We found the stage which was taking time. Refer details as below - This is simply reading file and mapping data to final save location. So, there are not much joins or calculations involved. Second, We saw event time line, if there is any delay due to serialization/ shuffling/ scheduling. But, there was nothing. Third, we saw total executors and tasks processed by them. There are 5 executors each taking over 3 hours and each executing approximately 500 tasks. Which means that a task is taking almost 2.8 -3 minutes Fourth, to confirm there is no data skew, we sorted the tasks to see maximum duration and maximum input a task processed, and maximum shuffle. And, we found nothing. So conclusively, we could say that there is no problem with Job. Why Job is slow, is because it has less number of executors or eventually less vcores for task processing. Thus, we bumped up Number of Executors from 5 to 30. And, each executo