Q&A

On Cloudera Public Cloud the storage unit is GCS. Therefore, hive tables and any inserted data are stored on GCS rather than local disks ( like as on-prem ). But, this adds an overhead because when running a job, the node manager needs to get the data from gcs and cache it locally till the job is done. This eventually hits the performance on the env. Is there any way to move the data location from GCS to Local disks on Public Cloud clusters? As written on the docs, datahub hdfs/disks spaces are temporary places but I would take this risk in favor of performance. @steven-matison would really appreciate your help on this question. Thanks in advance.

Hello team, We are facing issue and expecting your help here regarding Problem Statement :We create groups using in Cloudera user management using APIs.We can see those groups in Cloudera user management but not  could not find them as Ranger Groups. It seems groups are not being synced in Ranger.What are the possible solutions ? Any API is exist to sync users and groups?

Hi team, I have the error Faled running performance when i run the "Inspect network Performance" all the nodes of clouster are good except one. I see this error: [24/Feb/2023 09:19:19 -0600] 1547408 MainThread redactor INFO Started launcher: /opt/cloudera/cm-agent/service/perf/host_perf_diag.py input.json logs/result.json[24/Feb/2023 09:19:19 -0600] 1547408 MainThread redactor ERROR Redaction rules file doesn't exist, not redacting logs. file: redaction-rules.json, directory: /run/cloudera-scm-agent/process/1502-host-perf-inspector[24/Feb/2023 09:19:19 -0600] 1547408 MainThread redactor INFO Re-exec watcher: /opt/cloudera/cm-agent/bin/cm proc_watcher 1547410Traceback (most recent call last):File "/opt/cloudera/cm-agent/bin/cm", line 8, in <module>sys.exit(main())File "/opt/cloudera/cm-agent/lib/python3.8/site-packages/cmf/main.py", line 399, in mainroot(obj=argparse.Namespace())File "/opt/cloudera/cm-agent/lib/python3.8/site-packages/click/core.py", line 1137, in __call

I have a pretty simple setup.  We have two NiFi servers on different networks, and they need to transfer data between them.  I have site to site set up and working using HTTPS, but for some reason it seems like every single file transfer takes a minimum of 15 seconds. A 1 byte file will take barely over 15000ms, at a transfer rate of a couple hundred bytes per second, or a much larger file will transfer in just under 30 seconds, at a transfer rate of about 85MB/s.  I'm guessing the actual transfer of the file is happening quickly, but the transfer speed takes into account the 15 second handshake....or whetever the process is that is taking 15 seconds each time. I looked around and couldn't find anyone else experiencing quite the same thing. I have tried using RAW instead of HTTPS, but I haven't found a working setup for that yet--every time I try I end up with the "Magic Header not found" message. Here are some snippets of my config files, let me know

Greetings Cloudera Community!!!We have recently upgraded from Nifi 1.9.2 version to Nifi version 1.16.2 and we are getting the "Failed to index Provenance Events" Error, not all the time but it occurs occasionally. We have not faced such kind of error in nifi version 1.9.2. Can anyone please help us to figure out the possible cause of the occurrence of that error. Though it was not impacting the flow but still its occurrence is random.  Please find the error below in nifi-app.log.   Logs:2023-01-04 06:21:23,713 ERROR [Index Provenance Events-1] o.a.n.p.index.lucene.EventIndexTask Failed to index Provenance Eventsorg.apache.lucene.store.AlreadyClosedException: this IndexWriter is closedat org.apache.lucene.index.IndexWriter.ensureOpen(IndexWriter.java:877)at org.apache.lucene.index.IndexWriter.ensureOpen(IndexWriter.java:891)at org.apache.lucene.index.IndexWriter.updateDocuments(IndexWriter.java:1468)at org.apache.lucene.index.IndexWriter.addDocuments(IndexWrite

Hello, I have a Nifi Flow working to ingest logs messages and send it up to ElasticSearch. Everything worked fine for one month but suddenly some issues appeared without changes about size and volumetry of incoming logs messages: - I have a lookupRecord processor to enrich data and the queue become full for it so i guess this processor is the problem but it never had this behavior before- Nifi tell me it's unable to write flowfile content to content repo due to archive file size constraint- The input processor of the flow is a listenUDP and warn me that the internal queue is full I tried to reboot everything, obsioulsy, and also to increase JVM memory settings to Xmx12g to be sure it's not a lack of memory.I saw that the archive cleanup issue could be a bug from Nifi but I'm actually on 1.17 and even tried to move to 1.20 and everything is same. Any idea of what could cause those problems ?Thanks

We have recentely upgraded our CDP environments from runtime 7.2.12 to 7.2.16 and we are getting a wierd vertex error that we are unable to find the root cause for. anyone have a suggestion where to start looking? ERROR:Vertex vertex_1677501096465_0141_1_61 [Reducer 21] killed/failed due to:OTHER_VERTEX_FAILURE, counters=Counters: 0, vertexStats=firstTaskStartTime=1677588363293, firstTasksToStart=[ task_1677501096465_0141_1_61_000000 ], lastTaskFinishTime=1677591606664, lastTasksToFinish=[ task_1677501096465_0141_1_61_000000 ], minTaskDuration=-1, maxTaskDuration=-1, avgTaskDuration=-1.0, numSuccessfulTasks=0, shortestDurationTasks=[ ], longestDurationTasks=[ ], vertexTaskStats={numFailedTaskAttempts=0, numKilledTaskAttempts=2, numCompletedTasks=1, numSucceededTasks=0, numKilledTasks=1, numFailedTasks=0}, servicePluginInfo=ServicePluginInfo {containerLauncherName=TezYarn, taskSchedulerName=TezYarn, taskCommunicatorName=TezYarn, containerLauncherClassName=org.apache.tez.dag.ap

I'm using a tool in which I have to point out the master node (driver node) of the Cloudera Spark Cluster (spark :// <some-spark-master> : 7077). Also as I learned, Spark has "Master Node" (Driver Node) and "Worker Nodes".So I decided to go to the Cloudera Web Manager and checked the Configuration Tab of the Spark service, but all I found are "Gateway instance" and "History Server instance". Where are the "Driver instance" and "Worker instance"? I can't add these two instances in the "Add Role Instances" too My guess is that it's in Yarn service configuration, but I can't find anything related to "Master"/"Driver" or "Worker" either. So what is the link to "Spark Master" that ends with 7077? I can't find it anywhere in the Configuration tab 

When I start restart Master is OK.but after 10s Master is shutdown.I don't see error in log. I only find  CDH 6.3.2 enable kerberos

Hello i have the following problems  when i use hbase hbck 10 inconsistencies detected.Status: INCONSISTENT2023-02-01 13:51:01,346 INFO [main] zookeeper.ZooKeeper: Session: 0x1860d1d2ec00091 closed2023-02-01 13:51:01,346 INFO [main] client.ConnectionManager$HConnectionImplementation: Closing master protocol: MasterService2023-02-01 13:51:01,346 INFO [main] client.ConnectionManager$HConnectionImplementation: Closing zookeeper sessionid=0x1860d1d2ec000922023-02-01 13:51:01,346 INFO [main-EventThread] zookeeper.ClientCnxn: EventThread shut down2023-02-01 13:51:01,349 INFO [main] zookeeper.ZooKeeper: Session: 0x1860d1d2ec00092 closed2023-02-01 13:51:01,349 INFO [main-EventThread] zookeeper.ClientCnxn: EventThread shut down Also am unable to create a new table and this is the message that is being displayed  when i use create 'users','userinfo'ERROR: java.io.IOException: Table Namespace Manager not ready yet, try again later 

Hi all, I am new to Cloudera. Recently I am studying on CDSW and would like to have a try on it. May I ask is there any way that I can try CDSW for free? I don't have a paid account. I saw that there is 60 days trial on CDP Public/Private Cloud, is CDSW included? Hope someone can help me out. Thank you very much.

Hi, I have a workflow that is picking up an Excel file, that contains 3 sheets, and is attempting to run it through a ConvertExcelToCSVProcessor, but it is failing with the error below:  Failed to process incoming Excel document. Tried to allocate an array of length 328,219,733, but the maximum length for this record type is 100,000,000. If the file is not corrupt or large, please open an issue on bugzilla to request increasing the maximum allowable size for this record type. As a temporary workaround, consider setting a higher override value with IOUtils.setByteArrayMaxOverride(): org.apache.poi.util.RecordFormatException: Tried to allocate an array of length 328,219,733, but the maximum length for this record type is 100,000,000. If the file is not corrupt or large, please open an issue on bugzilla to request increasing the maximum allowable size for this record type. As a temporary workaround, consider setting a higher override value with IOUtils.setByteArrayMaxOverri

Question;We have around 6k tables under Hdfs as directories and when we use hdfslist processor in nifi its listing everything and killing the cluster . How can we select limited number of tables may 100 or 200 instead of listing everything ?

I followed https://docs.cloudera.com/cdp-private-cloud-base/7.1.8/securing-hive/topics/hive-enable-tls.html to configure Hive Kerberos with SSL. There was no issue, however , when I tried to obtain/display certs on client side to be used in connection, I obtained the following: openssl s_client -connect my_host:10000 -showcerts       CONNECTED(00000005)read:errno=0---no peer certificate available---No client certificate CA names sent---SSL handshake has read 0 bytes and written 287 bytes---New, (NONE), Cipher is (NONE)Secure Renegotiation IS NOT supportedCompression: NONEExpansion: NONENo ALPN negotiatedSSL-Session:    Protocol : TLSv1.3    Cipher : 0000    Session-ID:     Session-ID-ctx:     Master-Key:     Start Time: 1675956752    Timeout : 7200 (sec)    Verify return code: 0 (ok)

Hello,   I'm facing issue while executing query (show databases;) on hive.   Following is my setup: Cluster : CDP-7.1.7 Components Installed : HDFS, Yarn, Yarn Queue Manager, Oozie, Schema Registry, Hive, Ziplin, Spark, Tez and Zookeeper.    I have accessed hive using beeline, but when I tried to execute 'show databases' it is giving following error :   ERROR : FAILED: Hive Internal Error: org.apache.hadoop.hive.ql.metadata.HiveException(Error while invoking PostHook. hooks: java.lang.RuntimeException: java.lang.NullPointerExceptionat com.cloudera.navigator.audit.hive.HiveAuditPipeline.getInstance(HiveAuditPipeline.java:50)at com.cloudera.navigator.audit.hive.AbstractHiveExecHookContext.getPipeline(AbstractHiveExecHookContext.java:86)at com.cloudera.navigator.audit.hive.AbstractHiveExecHookContext.run(AbstractHiveExecHookContext.java:69)at org.apache.hadoop.hive.ql.HookRunner.invokeGeneralHook(HookRunner.java:296)at org.apache.hadoop.hive.ql.HookRunner.runPostE

Hello, I have a question about the performance of the local index in Apache Phoenix  For example, I have a table TBL_X having 6 columns (A , B ,C ,D E, F).1. Primary Key is (A, B) 2. One of the query patterns is   "Select * from  TBL_X where A = 'val1' and C = 'val3' "  For the better performance of the above query, I try to build the local index. Does   [ Local index on (C, A)  ]  perform better than   [ Local index on (C) ]  ?? or Will there be no performance difference since A is already in the first part of the PK ??   

I've installed and configured MIT Kerberos for Windows Hive ODBC connection.I could not create Kerberos ticket for more than one user, the new ticket created overwrites the existing ticket. Hence multiple user accessing that Windows machine could not obtain a valid ticket to connect to Hive ODBC connections.I've set krbs.conf from the Hive server and KRB5CCNAME is set to C:\temp\krb5cache.Please let me know how to configure this to have more valid tickets for each user.

table is `HiveFileFormat`. It doesn't match the specified format `ParquetFileFormat` I have this issue when I try to write using pyspark with the following command:df.write.mode("append").format("parquet").saveAsTable("schema.table") Before you say change from parquet to hive i know it works. But the thing is the table is partitioned in parquet and I really don't know why not its not working any more. It worked fine until now. The same command ran correctly for 1 month and 5 times so far. But today it does not want to write like this any more. If i check the metadata it also points to everything being in parquet:``103SerDe Library:      org.apache.hadoop.hive.ql.io.parquet.serde.ParquetHiveSerDeNULL104InputFormat:        org.apache.hadoop.hive.ql.io.parquet.MapredParquetInputFormatNULL105OutputFormat:       org.apache.hadoop.hive.ql.io.parquet.MapredParquetOutput

Hi All, I am currently building a pipeline that fetches all changes from 2 mysql DBs into a final mysql warehouse. I am using changecapturemysql processor to identify inserts,updates and deletes and then sending them to putdatabaserecord processor. All inserts work good. My source has a PK (primary key) column. My destination has the same column + another column as PK . I am able to replicate inserts. But when it comes to updates and deletes,  updates on non primary keys were getting captured, but if i update primary key column in source, it is unchanged in destination. same with deletes. a delete with where clause on PK, doesnot deleting the record in destination. As i read the documentation this is expected behaviour - "MPORTANT: If the Statement Type is UPDATE, then the incoming records must not alter the value(s) of the primary keys (or user-specified Update Keys). If such records are encountered, the UPDATE statement issued to the database may do nothing (if no exis

We have a SQL that normally runs for 2 mins including a few stages, including 8 mappers and 1000 reducers. One day, among the 8 mappers, 7 finished in <1 min but 1 stuck for 2.5 hours. The logs in that container stopped after 1-2 mins (the task was killed so I could not produce the logs but there were nothing suspicious). The counters show that there were no activities at all for a very long time, CPU usage low, network, IO, all low. The 1000 reducers were up and waiting, using all the resources and kept the rest of the tasks waiting. Re-running the SQL with the same data (no skew) multiple times and it always completes within 1-2mins. What could cause a mapper to stuck silently? Is there a way to debug this (the logs were not very helpful)? This happens occasionally for different SQLs, is there a way or tool to detect such situation and restart / kill the job? Thanks for the help!

How to upload files to cloudera virtual box? I'm taking SimpliLearn's Big Data Hadoop and Spark Developer Training course, which is a free course. That's why I don't have access to the labs. So what I did was download cloudera virtual box to practice spark and hadoop codes. However, some exercises require uploading files such as datasets using FTP.I tried using FileZilla to connect to the VM and upload the files to the VM, but it didn't work connecting using the VM's IP address.I'm a beginner with these tools, could anyone help me with how to put these files in the VM? See below what I need to do:1. Download the Hadoop-mapreduce-example.jar file and wordcount.txt file2. Log in to the FTP using the username and password from the lab and upload the file3. Log in to the Webconsole using the username and password from the lab and create a new directory demo in HDFS using the mkdir command4. Push the wordcount.txt file into the directory using the put command5. Execute the command

How to run selected processor group in apache nifi while restarting nifi?

How to set the partition life cycle of the hive table？ Error: Error while compiling statement: FAILED: ParseException line 1:270 missing EOF at 'lifecycle' near ')' (state=42000,code=40000)

Hi Team,We have CDP cluster integrated with OpenShift cluster, private cloud setup with all the service like hive, atlas, ranger running. When we are creating the Datawarehouse CatLog in management console we are getting the below error in console, can you please help us to find out the cause.some Errors from management console: 1. Driver failed to create DB Catalog: IP address is not expected for default warheouse RDS endpoint. Check if RDS is enabled in the env? Error Code : 99992. timeout exceeded transitioning from state 'Starting'.Last event: {InProgress checking if metastore statefulset is ready with at least 1 ready replica(s) (config-id: 5afc7144-9409-47e2-a9b9-2c7822adafbf version: 2022.0.7.1-2) 2023-02-09 10:08:44.366274685 +0000 UTC} Error Code : 9999OpenShift pods log:/usr/local/bin/config-merger /tmp/hive/conf/atlas-application.properties /mnt/config/current/atlas-application.properties /etc/hive/conf/atlas-application.properties -wf2 atlas.rest.address -wf2

To get started with CDP public cloud we have to create a env that is named as datalake and if we wanna run any jobs/workload on top of it we have to create datahubs based on the services we need.  Inside the datahub cluster, there is an hdfs service that serves as intermediate place to store data while working on to be eventually stored/moved to cloud object storage. But there is another hdfs service inside the dataleke? what is usage of it? it's total capacity of 1.4TB SSD, how can we utilize it?