Prod - DRP ( disaster recovery plan )

[LaurentMag]

Summary of the process general steps :

• Determine what kind of incident to cover (DDOS / Database issues / Network failures etc... )
 recovery strategies, resources required to recover, predetermined and acceptable timeframe



RTOs (recovery time objective)
 RPOs (recovery point objective)

To minimize interruptions to normal operations.


To limit the extent of disruption and damage.
To establish alternative means of operation in advance.


• Determine what is critical, and the most important ( application / data ), determine what to save and what can’t be delayed (depending of the application) here

  • Docker images
  • Workflow to build everything ( is there workflow on the prod server to save or recover from.. )
  • The data from the database

• Determine who will handle, and which part of the procedure

• Test the DRP

• Update the DRP

[LaurentMag]

What's decided for the Yaki DRP :

• wipe the entier prod server, it must be clear as if its a new server
• Need to put back on the server all the necessary configurations files
• Recreate the docker based on dockerfiles/docker compose
• set database data back

This will be done manually for now

[LaurentMag]

Must consider differents situations :

• OVH services are down
• Need to create a new prod (create a new folder, consider to have a script setting (.env ngnix.conf docker compose)