Disaster Recovery Procedure for UniField Servers

Disaster Recovery Procedure for UniField Servers

This document provides a framework of planning for the risks of server outages. It outlines the steps and systems for preventing server disruptions, as well as protocols for responding to outages and recovering lost data. Below are the main objectives of the document.

1) Objectives

• • To help OCs develop or integrate this procedure to their Disaster Recovery Plan in order to adequately prepare for an unforeseen disaster
  • Help ensure rapid recovery after a disaster that has impacted Unifield servers, thus minimising impact on field operational activities.
  • Provide instructions, procedures, and emergency contact information to use in a disaster situation
  • Identify processes to follow to ensure server restoration after a critical event.
  • Identify current risks and recommend action steps for prevention

2) Points of Communication

Fire, theft, floods, botched system upgrades or simply human error: any of these could take down your unifield server. If this happens, do not attempt to restore the instance on your own but immediately create a jira ticket or send an email to the following points of contact in the unifield coreteam.

Name	Title	Email	Specific Roles and Responsibility in Disaster Recovery
Raffaelle HAGEN	Head of Support and Development	raffaelle.hagen@geneva.msf.org	Initiates, validates, and oversees the disaster recovery process
Awfa AbdulGhany	ERP Support Officer	awfa.abdulghany@sits.msf.org	Implements server recovery procedure by: -Determining the appropriate recovery point from backups. -Liaises with dev to check backups for integrity and initiates data recovery processes for lost data.
Rafkat Iskakov	ERP Supply Support Manager	rafkat.iskakov@brussels.msf.org	-Ensures that all supply users are notified of the incident through their respective supply referents. -Validates supply recovery points for cases where supply data is lost and helps define next steps for business continuity
Estibaliz Montaru	ERP Finance Support Manager	estibaliz.montaru@geneva.msf.org	-Ensures that all finance users are notified of the incident through their respective finance referents. -Validates finance recovery points for cases where data is lost and helps define next steps for business continuity

3) Response

Do not restore the instance from the last backup you have. Instead, create a jira ticket or send an email to raffaelle.hagen@geneva.msf.org and awfa.abdulghany@sits.msf.org . Please include the following information:

• • Description of what happened to the server
  • Share the date of the last updated local backup you have.

Actions/Steps we will take after we receive your ticket or message.

Please note that the following actions are in chronological order:

a) Disable sync for the instance on the sync server.

b) Determine the appropriate recovery point from backups by:

• • Comparing the latest copy of the instance backup we have on our continuous backup server to the last local backup you have. If the backup dump on our continuous backup server is more updated than the local dump you have, we will provide you with the dump and give you the green light to restore the instance. Do not let users use the server until we give the greenlight.
  • If the local backup dump you have is more updated than the one on our continuous backup server, we will request for a copy of the local backup dump you have. Wait for the greenlight from us as we check the dump for integrity.

c) Check backups for integrity and initiate data recovery processes for lost data.

• • Users will not be allowed to access or use the instance until the data recovery processes are completed.
  • The Finance and Supply ERP Managers will communicate to the respective OC referents the non-synched data that has been lost (cannot be recovered as it was not pushed to the sync server). Together with the OC referents, they will determine next steps for business continuity.
  • A data fix will then be prepared for the lost data that was synched to the sync server (data re-created from the sync server).
  • We will then apply the data fix to the restored dump, the coordination instance and HQ instance plus any other impacted project instance. Wait for the official green light from us before allowing users to access and resume working on the instance.

4) Preventative & Recommended Guidance

Disasters happen, some unexpected like fires, theft, floods but others we can anticipate. The only preventative and recommended guidance we advise is backup! Data backup is the foundation of disaster recovery planning.

Backup Strategy

As good practice do not keep the local backup of your instance on the server. The continuous backup feature ensures UniField pushes backups each time an instance synchronises. These are, however, dependent on internet connection. Please DO NOT CONSIDER this feature as an agreed offsite backup solution. Backups are still under the responsibility of each OC. Therefore, each OC ought to be vigilant with proper backup strategy and solutions to their servers.

We also recommend that you:

• • Backup your instance before any system upgrades.
  • Backup the instance before upgrading or making changes to any third-party applications on the server.
  • Backup before and after migration of an instance.
  • If in doubt, just back up the instance