Hello @SAMSAL ,   thank you for reaching out with your issue about parsing a CSV with SplitRecord processor. The described behaviour sounds like an encoding issue. Have you tried to identify what is the encoding of your source data, please? I've just quickly googled for "identify encoding" and found this tool, which might works. (Have not tested it, so feel free to browse in this topic).   Once you know what is the character encoding for the source data, please set your "Character Set" in NiFi accordingly.   Please let us know if it helped by pressing the "Accept as Solution" button.   Kind regards: Ferenc ... View more

Hello @Jua ,   thank you for reporting the issue of CM not being able to connect to it's embedded database after installing manually postgres client rpm.  I understand that you already tried to start the scm server db.    Do you still face the issue, please or was it resolved since?   Please verify if: the postgres database is running and can be accessed? the db is running... once you kill the db process and restart the cloudera-scm-server-db.service, does it fix the issue?   Thank you: Ferenc   ... View more

Hello @gss2020 ,   thank you for exploring our HDP Sandbox. For the default passwords, please see our "Learning the Ropes of the HDP Sandbox" tutorial page. Within this, please have a look at "Appendix A: Reference Sheet" for the login credentials.   Please let me know if you need any further input from us.   Thank you: Ferenc ... View more

Hello @Nagamalleswara ,   please make sure you are using the yarn user when collecting the thread dumps.   I've found useful this description too for figuring out what is going on with the JVM.   Kind regards: Ferenc ... View more

Hello @Nagamalleswara ,   What is odd is that the issue happens on large AWS nodes only and intermittently. Did you check if ports are open bi-directionally on the affected nodes for NMs? Can you access the NM WebUI on the affected node, please?   Thank you: Ferenc ... View more

Hello @netapp1 ,   thank you for your deep dive into the CDH6.x clusters CM Express version's 100 nodes limitation.   The intent from Cloudera is to limit Cloudera Express 6.x usage to 100 nodes. The implementation on how it was enforced from a User Experience point of view is different between minor versions (e.g. 6.0. vs 6.1. vs 6.2.). It means that earlier versions might let you install the agents across the nodes and once the installation completes it's blocks most of the functionalities until you decrease the number of nodes to 100.   We recognised that it would be best to block installation until the nodes are not exceeding 100 nodes and in later version it's been gradually improved.   Hence the discrepancy between the descriptions in the documentation.   In short: 100 node limitation is enforced across Cloudera Express 6.x versions.   Please let us know if we addressed your enquiry on this topic by pressing the "Accept as Solution" button, which helps other members to find the answer of a similar question.   Thank you: Feren ... View more

Hello @Nagamalleswara ,   thank you for sharing with us the detailed description of how your NM seems to have issues on a certain node.   Next time you hit this problem, can you please check if you see multiple NM instances on a single node? In some cases the CM agent might gets into a bad state.   When you hit the issue, please run the following command and see if it showes 2 NM processes running on the host ssh into the host ;   become root   # export SUPER_CONF=/var/run/cloudera-scm-agent/supervisor/supervisord.conf # /usr/lib64/cmf/agent/build/env/bin/supervisorctl -c $SUPER_CONF status 52372-yarn-NODEMANAGER       RUNNING    pid 71771, uptime 2:21:19 52723-yarn-NODEMANAGER       FATAL      Exited too quickly (process log may have details)   If so, it means that the CM-agent process was bad.   In this case please: 1) Proceed to kill the PID. Verify that the PID does not exist and neither is anything else listening on the port. 2) Restart cloudera-scm-agent  service cloudera-scm-agent restart 3) Start the RM process from CM   Please share with us your findings. If you are not hitting the above described issue it still helps us to narrow down further this issue.   Please also check if the NM heap size on this node is set to at least 1Gb just to rule out stop the world GC activity rendering the NM unresponsive.   Thank you: Ferenc ... View more

Hello @getschwifty ,   did you have a chance to test if cross-realm trust works OK between your nodes from different realms?   Kind regards: Ferenc ... View more

Hello @getschwifty ,   thank you for enquiring about how to set up RPC HDFS remote access. Please see this blog article about WebHDFS that should do the job.   Please let us know if we addressed your enquiry by pressing the "Accept as Solution" button. It also helps search engines to surface the answers that were actually helpful.   Best regards: Ferenc ... View more