Symptoms

  1. After the restart host with Instance Manager Database, Instance Manager service PACI-im does not start/restart on Instance Manager node.

  2. There are no errors in /var/log/messages on Instance Manager node.

  3. You detect the errors related to the database schema in /var/log/pa/vps.log on Instance Manager node similar to the shown below:

    2017-04-13 20:48:15,847 () INFO  ValidateMappers [main] - Validating mapper: com.parallels.c2u.im.mappers.BackupNodeMapper
    2017-04-13 20:48:16,212 () ERROR ValidateMappers [main] - Invalid SQL statement found at com.parallels.c2u.im.mappers.BackupNodeMapper.addNode:
    INSERT INTO backup_node (uuid, name, ip, login, password, pdata, capacity, for_backups, for_images, state, location, virtualization_technology) VALUES(?, ?, CAST(? AS inet), ?, ?, ?, ?, ?, ?, ?, ?, ?)
    ERROR: column "virtualization_technology" of relation "backup_node" does not exist
      Position: 118
    
    2017-04-13 20:48:16,220 () ERROR DataSourceModule [main] - Internal validation failed. Terminating.
    org.postgresql.util.PSQLException: ERROR: column "virtualization_technology" of relation "backup_node" does not exist
      Position: 118
    

Cause

PostgreSQL 9.1 is started instead of PostgreSQL 9.6. This is a result of bug CCU-14831:

Auto-start of service postgresql-9.1 is not disabled on IM DB node after OA upgrade from 7.0.1 to 7.1

Resolution

To resolve the issue, do the following on IM DB node:

  1. Check runlevels of postgresql-9.1 and postgresql-9.6 services:

    # chkconfig --list postgresql-9.1
    postgresql-9.1  0:off   1:off   2:on    3:on    4:on    5:on    6:off
    
    # chkconfig --list postgresql-9.6
    postgresql-9.6  0:off   1:off   2:on    3:on    4:on    5:on    6:off
    
  2. On IM DB node, configure stopping PostgreSQL 9.1 automatically during the system startup by executing the command:

    # chkconfig postgresql-9.1 off
    
  3. Stop PostgreSQL 9.1:

    # service postgresql-9.1 stop
    
  4. Start PostgreSQL 9.6:

    # service postgresql-9.6 start
    
  5. On IM node, restart the service PACI-im:

    # service PACI-im restart
    

Internal content