• vIDM DB partition is full or nearly out of space.
• To validate which DB partition is using most of the space run the following commands:

  du -sh /db/elasticsearch
  du -sh /db/rabbitmq
  du -sh /db/data

• The cause is over utilization of API by other clients. 
• Any time the API is invoked it creates a refresh token entry in the database utilizing additional storage.

Prerequisites

• You have root access to the appliance.
• You have valid backups or snapshots.

Steps to clean up excessive disk space consumed by OAuth2RefreshToken table.

1. SSH into the primary appliance where the Master role for vPostgres is located
2. Stop the horizon-workspace service on all nodes

   service horizon-workspace stop

3. Find the largest table sizes

   export PGPASSWORD=`cat /usr/local/horizon/conf/db.pwd`
   /opt/vmware/vpostgres/current/bin/psql -U postgres saas
   
   SELECT nspname || '.' || relname AS "relation",
   pg_size_pretty(pg_total_relation_size(C.oid)) AS "total_size"
   FROM pg_class C
   LEFT JOIN pg_namespace N ON (N.oid = C.relnamespace)
   WHERE nspname NOT IN ('pg_catalog', 'information_schema')
   AND C.relkind <> 'i'
   AND nspname !~ '^pg_toast'
   ORDER BY pg_total_relation_size(C.oid) DESC
   LIMIT 20;

4. Remove the data from the table by running a truncate

   truncate table "OAuth2RefreshToken" CASCADE;
   

5. Reindex the database and vacuum

   select count (*) from "OAuth2RefreshToken";
   select count (*) from "OAuth2AccessToken";
   select count (*) from "SuiteTokenCache";
   vacuum full verbose;
   reindex table "OAuth2RefreshToken";
   vacuum full verbose;
   \q

6. Restart the vPostgres service

   service vpostgres restart

7. Start the horizon-workspace service on all nodes

   service horizon-workspace start

8. Wait for 10 minutes after restart to allow services to fully initialize.
9. Count the number of rows in the OAuth2RefreshToken table

   /opt/vmware/vpostgres/current/bin/psql -U postgres saas
   select count (*) from "OAuth2RefreshToken"