Skip to main content
Ben1996
Ben1996
Silver ⭐️⭐️
June 10, 2024
Solved

Backup Strategy for CDP Private Cloud

We are on CDP 7.1.7 SP2 version

Can I ask what is the best strategy to backup the cluster? We are using 5 phyisical nodes (3 of which holds the actual HDFS data). What is a good strategy for backup?

1. Create another cluster that will act as replica (lower CPU and RAM)

2. In the same cluster, backup / to preserve the OS apps (CM etc.) and /data (HDFS)

    Rajat_710
    Rajat_710Answer
    Gold ⭐️⭐️
    June 12, 2024

    hey, @Ben1996 what would like to backup is the HDFS data or anything else?

    Also, I would like to let you 1 thing if possible add 1 more data node in your cluster as by default you have 3 as RF and if 1 DN goes down then you'll have only 2 DN alive which could be critical for business.

    For backing up HDFS data you can create anther cluster with low resources and setup a BDR jobs on it to replicate data from PROD to DR cluster refer to the below link.

     

    https://docs.cloudera.com/cdp-private-cloud-upgrade/latest/release-guide/topics/cdpdc-compatibility-matrix-bdr.html

    View in thread

    5 replies

    Rajat_710
    Gold ⭐️⭐️
    Rajat_710Answer
    Gold ⭐️⭐️
    June 12, 2024

    hey, @Ben1996 what would like to backup is the HDFS data or anything else?

    Also, I would like to let you 1 thing if possible add 1 more data node in your cluster as by default you have 3 as RF and if 1 DN goes down then you'll have only 2 DN alive which could be critical for business.

    For backing up HDFS data you can create anther cluster with low resources and setup a BDR jobs on it to replicate data from PROD to DR cluster refer to the below link.

     

    https://docs.cloudera.com/cdp-private-cloud-upgrade/latest/release-guide/topics/cdpdc-compatibility-matrix-bdr.html

    Translate
      Ben1996
      Ben1996Author
      Silver ⭐️⭐️
      June 13, 2024

      Thanks Rajat! Does that mean if 1 failed, the entire cluster will not be functional? Thank you for linking the replication manager, so it is essentially option 1 of what i previously thought of.

      Translate
      Rajat_710
      Gold ⭐️⭐️
      Rajat_710
      Gold ⭐️⭐️
      July 16, 2024

      no @Ben1996 if 1 DN fails it won't crash the cluster but you'll be more vulnerable and run a cluster with only 2 copies of data in DNs which is not recommended for the PROD environment.

      Translate
        VidyaSargur
        Community Manager
        VidyaSargur
        Community Manager
        July 22, 2024

        @Ben1996, Did the response assist in resolving your query? If it did, kindly mark the relevant reply as the solution, as it will aid others in locating the answer more easily in the future.

        Translate
        Regards, Vidya Sargur, Community Manager Was your question answered? Make sure to mark the answer as the accepted solution. If you find a reply useful, say thanks by clicking on the thumbs up button.
        Ben1996
        Ben1996Author
        Silver ⭐️⭐️
        July 22, 2024

        Thanks @Rajat_710  and @VidyaSargur  . I already accepted the answer.

        Translate
          Terms & ConditionsAccessibility statement