Step-1:
Get the nightly backup from the customer's Production Concerto node:
/var/versa/ecp/archives/
Sample:
admin@versa-ecp:/var/versa/ecp/archives$ ll
total 2889672
drwxr-xr-x 2 999 versa 4096 Sep 30 00:03 ./
drwxrwxr-x 7 versa versa 4096 Aug 3 14:12 ../
-rwxr-xr-x 1 999 versa 2059533167 Sep 30 00:03 builds.tgz*
-rwxr-xr-x 1 999 versa 43035027 Sep 29 00:06 dbexport_2024-09-29_00-00-15_concerto-20240927-142607-3371e92-12.2.1_Nightly.pgsql.tgz*
-rwxr-xr-x 1 999 versa 43143114 Sep 30 00:03 dbexport_2024-09-30_00-00-10_concerto-20240927-142607-3371e92-12.2.1_Nightly.pgsql.tgz* >>> For example
Step-2:
SCP the, psql-database to the Lab Concerto running the same version under /var/versa/ecp/archives/
Step-3:
Hashout the "--migrate-deploy" line in /opt/versa/ecp/scripts/utils/pg_backup.sh
Sample:
Step-4:
Take the backup of the /var/versa/ecp/share/config/clusterconfig.json file.
Step-5:
Load the backup using this command.
vsh database backup restore <Backup-file-name>
Sample:
admin@versa-ecp:~$ vsh database backup restore dbexport_2024-09-30_00-00-10_concerto-20240927-142607-3371e92-12.2.1_Nightly.pgsql.tgz
Step-6:
The backup may fail during startup, hit Ctrl+C to terminate.
Copy back the cluster_config.json that was backed up to /var/versa/ecp/share/config/ from Step-4.
Run:
vsh cluster regenerate-manifest
Step-7:
Update the Director key_value to a internal IP which would be the Public WAN IP of the production directors.
Sample reference:
admin@versa-ecp$ pgdb
portal=# select * from director_metadata ;
uuid | key_name | key_value | director_cluster_uuid
--------------------------------------+--------------+---------------+--------------------------------------
20416f09-1595-4be7-aa57-2044114b7bed | DIRECTOR_IP1 | 10.48.28.177 | f5795a26-e6e5-4665-9f45-7043edfd42b5
f018551c-1cf8-4054-bad8-6c76294a792b | DIRECTOR_IP2 | 10.48.28.178 | f5795a26-e6e5-4665-9f45-7043edfd42b5
e74f87eb-9e7e-496d-8c41-595a26725789 | DIRECTOR_IP1 | 10.40.199.227 | 8c46e969-773e-4646-b83f-ca1882874e1c
(3 rows)
We need to update all the director_metadata to point to an internal dummy IP.
portal=# update director_metadata set key_value='10.40.1.1' where uuid ='e74f87eb-9e7e-496d-8c41-595a26725789';
portal=# update director_metadata set key_value='10.40.1.2' where uuid ='f018551c-1cf8-4054-bad8-6c76294a792b';
portal=# update director_metadata set key_value='10.40.1.3' where uuid ='20416f09-1595-4be7-aa57-2044114b7bed';
Step-8:
Update the analytics_cluster and analytics_aggregator, north_bound_ip and ip_list respectively.
Sample:
portal=# select * from analytics_cluster ;
uuid | name | north_bound_ip | director_cluster_uuid | is_default | role
--------------------------------------+-------------+----------------+--------------------------------------+------------+---------
a1cc5b03-c305-49b0-ad7a-833095696c79 | VAN-Cluster | 10.48.28.175 | f5795a26-e6e5-4665-9f45-7043edfd42b5 | t | UNKNOWN
portal=# select * from analytics_aggregator ;
uuid | name | ip_list
------+------+---------
(0 rows)
>>>>> Update the north_bound_ip and ip_list respectively.
Sample step to update the IP's:
portal=# update analytics_cluster set north_bound_ip='10.40.1.4' where uuid ='a1cc5b03-c305-49b0-ad7a-833095696c79';
If there is an analytics_aggregator, then use this command to update the aggregator IPs.
portal=# update analytics_aggregator set ip_list='10.40.1.4' where uuid ='aweweb03-c305-49b0-ad7a-8330956wewe2';
Step-9:
Perform "vsh start".
+++++++++++++++++++++++++++++++++++++++++++++++
How to reset the UI admin password for users to be able to login?
admin@versa-ecp:~$ pgdb
portal=# update login_user set password='$2a$10$rZfVj2Hfv38MEpJLXqyqReog86sFYtdvPViXgxH5rOvuDoj0v2xD6' where username='admin';
portal=# exit
This way we should be able to Login to the UI using username: admin Password: versa123
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
FAQ:
1) How to list the backups present in Concerto?
admin@versa-ecp:~$ vsh database backup list
2) Raised an internal Feature request to provide a vsh command to be able to perform this steps.
Bug-ID: 116192
Description: Provide a vsh command to safely load customer backups internally