Recovery Procedures

This chapter explains how to recover CloudCIX functionality from the PAT in the event of a failure.

Podnet A Failover to B v5.0.0

1. Move the “PodNet Public” cable from Podnet A and connect to the same port on Podnet B.

2. Connect keyboard and mouse to PodNet A.

   • If PodNet A is responsive, run:

     bash /etc/cloudcix/pod/pod_installer/ha_scripts/disable.sh
     

   • If PodNet A is unresponsive, disconnect the power to PodNet A.

3. Connect keyboard and mouse to PodNet B and run:

bash /etc/cloudcix/pod/pod_installer/ha_scripts/enable.sh

Podnet Rebuild v5.0.0

Podnet v3.0.0

1. Get “region_id” from the Membership App, open “My Membership” > “Addresses”. The “region_id” is the “Address ID”.

2. From PAT VDI, open pgAdmin (http://[<*prefix>:d0c6::5002:1]) and open connection to the pgsqlapi database

3. Open “Query Tool” for “iaas” database.

-- Get a list of Projects in the Region using the region_id from Step 1
SELECT * FROM project
  WHERE region_id = 3 AND closed = FALSE and deleted IS NULL;

-- Takse note of the `id` field in response as project_ids
-- Update Running virtual_routers to Requested state to rebuild
UPDATE virtual_router
  SET state = 1
  WHERE state = 4 AND project_id in (<project_ids>) AND deleted IS NULL;

4. Connect to POD Appliance for COP of the Region - <prefix>::6000:1 and run the following

cd /etc/cloudcix/docker/

sudo docker compose down iaas && sudo docker compose up -d iaas && sudo docker compose restart nginx1

5. Robot will start builidng the virtual routers in the region.

*<prefix> represents the first three hextets in the /48 IPv6 Addresses assigned to a Pod.

KVM Hosts

When a KVM host is powered on, the following steps need to be performed to prepare the host for use:

1. Mount network drives by running the following command:

sudo mount -a