Expert continuity tips from industry-leading team that supports more than 700 invocations a year.
We have consulted our industry-leading recovery engineers to extract some expert tips from the team that supports more than 700 invocations, standbys and rehearsals every year. These are well worth a read, to help ensure your business continuity services will be effective when you need them:
- If you make any business system changes, upgrades or add new applications please check your backup schedules so that you know these are going to be backed up
- Keep on top of your workplace recovery images, keeping them as up to date as possible
- If you make a systems change and go through CAB (change advisory board) then ensure your change includes business continuity, otherwise you may run into two problems: Firstly – can you go back? Is the change recoverable using your existing DR? And secondly – will this change break your BC plan?
- Make sure there are no single points of failure in your staff when a recovery is needed – for example, does more than one person know how to recover (each of) your systems?
- Don’t assume that “work from home with a laptop” is the answer to your BC problems. If you have to evacuate your building during business hours, how will this work in practice? Can everyone connect to the right systems? Are they on a secure network? How is telephony managed? How will staff collaborate? (The list goes on.)
- Ensure your staff are familiar with your work area recovery facility (where it is, local procedures, local food places etc). Bring them for a tour of the facility so they know what to expect if they have to work there
- Regularly review your authorised invoker list to ensure there are no delays when invoking your BC service
- Regularly review your Battlebox access list and your data centre access list – you don’t know when you will need it and you won’t want to be delayed with security when you do!
- Clients often change their equipment in their live system, but do not update their disaster recovery contract – this means that their equipment will not be ready on rehearsal or invocation which puts them at risk
- Clients sometimes book a rehearsal without fully completing a rehearsal booking form, then find that the equipment is not what they wanted to test or that there are not sufficient resources available – make sure you complete the form fully so that everything you want to test is covered
- Always backup the system state – without the system state, the recovered machine will probably not boot up afterwards. The system state is required for the server to be ‘as like’ the original server as possible.
- A handy tip is that when you upgrade a server, keep the pre-test information questionnaires up-to-date in-case of an invocation. We have all been witness to customers bringing in their replicated backup solutions, only to find that they are either not backing up complete backups, or backing up corruption
- It is important to validate your backups! There have been several instances where we’ve had issues with recoveries because the backups are not good or complete and we’ve been trying to recover a partly failed backup!
- We have instances where customers will tell us that, “nothing’s changed since last rehearsal” – when actually almost everything has changed! – It’s so important to fully complete the information required and to make sure it becomes part of your process to update the information when things change, so that there are no delays when it comes to recovery! Accuracy of pre-test information provided to us is often an issue where the customer does not complete them fully with correct OS versions, service pack levels, tape drive e.g. LTO, memory and hardware platform – we have to start recoveries with the correct OS loaded for example. If it’s not, then the recovery will fail and we then have to re-install the machine with the correct version, which obviously wastes time.
- Consider tape contention when backing up – some customers put all their big servers on the same tapes, which leads to the restores queuing whilst waiting for the tapes
- Ensure you have access to a record of all administrator passwords (local and domain) and any encryption keys used
- When they invoke/test, customers need to remember to bring the license files for their backup products with them (VEEAM, Netbackup etc.).
- Customers need to consider the catalog backup location when designing their backup solution. Having the catalog saves time for recoveries, because otherwise each tape will need to be scanned. Make sure the catalog is available and is not on-site. Some backup packages may send the DR file to an email address, so a cloud-based email address is a good idea as it can be retrieved from anywhere.
- The customer may replicate their environment or use some sort of cloud based backup, but an awful lot also have a final backup to tape. If this is you, consider testing the tape backups once in a while as you never know when you may need them!
- Test Test Test! Practice makes perfect
If you have any questions or would like further information about any of the above points, please email the Daisy recovery team on BCOpsAdmin@daisygroup.com