Restoring Data Visualization

You can restore the Data Visualization instance that you backed up.

Do not open the Data Visualization web interface prior to applying the steps below.

Find the necessary information like database name, host, port, user, and password. On the new cluster, this information will be different from the old cluster.

$ kubectl get secrets/pg-db-secret -o=jsonpath={.data.'\.pgpass'} -n <viz-id> | base64 -D

postgres-service:<port>:<database name>:<user>:<password>

For example:

$ kubectl get secrets/pg-db-secret -o=jsonpath={.data.'\.pgpass'} -n viz-1680129861-kbtr | base64 -D

postgres-service:5432:metastore:hive:ZufGC6Dmh03N1iW042uTosZtr4XvCtJIYPQ==

Using these properties you will be able to connect to the database and load the backup.

Use kubectl or k9s to access Hue container in the targeted CDW environment (need KUBECONFIG setup), and find the namespace for Hue and for Data Visualization.
```
$ kubectl get pods --all-namespaces --field-selector metadata.name=huebackend-0

$ kubectl get pods --all-namespaces --field-selector metadata.name=viz-webapp-0
```
If you have multiple Data Visualization instances running, match the new Data Visualization namespace names to the user-friendly names on the CDW UI.

Select one of the Hue namespaces and copy the dump file to your container.

$ kubectl cp ~/Downloads/logs/viz_pg_dump.tar <Hue container>/huebackend-0:/opt/hive/viz_pg_dump.tar -c hue

For example:

$ kubectl cp ~/Downloads/logs/viz_pg_dump.tar impala-1679934278-6pgc/huebackend-0:/opt/hive/viz_pg_dump.tar -c hue

Shell into the container.

$ kubectl exec -it huebackend-0 -n <Hue container> -c hue -- /bin/bash

Load the dump back to the database, taking care if you have multiple DataViz instances to load the contents back to the right database.

pg_restore -d <DataViz database ID> -h postgres-service ./viz_pg_dump.tar -c -U hive

For example:

pg_restore -d viz-1680137534-87s4_vizdb -h postgres-service ./viz_pg_dump.tar -c -U hive

The pg_restore command might output the following errors as it runs:

....
pg_restore: [archiver (db)] Error from TOC entry 258; 1259 302230 TABLE apps_apikey hive
pg_restore: [archiver (db)] could not execute query: ERROR:  table "apps_apikey" does not exist
    Command was: DROP TABLE public.apps_apikey;

WARNING: errors ignored on restore: 10

It is safe to continue restoring Data Visualization; just ignore the errors.