- Subscribe to RSS Feed
- Mark as New
- Mark as Read
- Bookmark
- Subscribe
- Printer Friendly Page
- Report Inappropriate Content
Created on 11-01-2018 09:02 PM - edited 09-16-2022 01:44 AM
Below are some FAQ's which helps you to quickly identify some important info for DPS-DLM deployment.
Pre-req's for DPS and DLM
DB version- postgres 9.3 to 9.6 OS - RHEL 7.0 and above Ambari - 2.6.2 HDP 2.6.5 Distcp should work b/w source and target clusters. Beacon user should be created in AD there is no choice of using custom user in this DLM 1.1 release the onboarded service user for your application should exists in AD and needs to be resolved(id <username> on both source and target clusters . docker version Required ports needs to be open b/w source and target clusters and also to access DPS UI
where to install DLM and DPS software components?
DLM engine needs to be installed as m-pack on both clusters using Ambari server DLM app is dockerized container needs to be installed on DPS host
Which URL needs to be given to register cluster in DPS UI
the Ambari URL integrated with knox http://<>:8443
I'm unable to see the DLM icon in DPS UI after enabling DLM component in DPS
User needs to be part of Infra-admin role
Verify DLM Engine install
Verify that Beacon was added as a user to the HDFS superuser group. hdfs groups beacon The output should display HDFS (or value of the dfs.permissions.superusergroup config) as one of the groups. Beacon user should be part of ranger policies https://docs.hortonworks.com/HDPDocuments/DLM1/DLM-1.2.0/installation/content/dlm_verify_the_dlm_eng...
Mostly used commands for for troubleshooting
On DPS host use below commands docker ps -- check ports,containers and uptime docker images docker exec -it <docker-name> docker exec -it 029ec380bb3d /bin/ls -alrt /usr/dp-app/ docker logs --follow dp-app docker exec -it d6390b6c0c50 /bin/ls -alrt /usr/dp-app/
Required Machine config for DPS and DLM
DPS runs on separate machine which will run all docker containers. <br>Master Node config is recommended for this host with atleast 64 GB of memory if you are using external database for same host consider more memory and CPU
For hive replication the in target cluster beacon is auto creating deny policy in ranger ..is this expected behavior or bug in DLM 1.1?
This is to prevent any writes from happening outside of replication to the target database the deny policy is only on the replication target database
For Hive replication can we schedule job per table basis?
No ,in this current DLM 1.1 release only database level is supported.
Please upvote if its helpful.