If you're using Backup Exec™ software, it's likely that you're using it to be prepared for disaster. If so, you'll want to ensure that if disaster strikes you're properly prepared.
You need to consider many aspects, including:
- Different types of disaster
- Making sure you can recover all your data
- Knowing what you need to do if disaster strikes
You simply can't consider every type of possible disaster. As a result you might want to consider the worst, and assume you might lose your entire production site. This would include the loss of all hardware and data stored on site. If you account for this, then you should probably be able to cope with any lesser disaster.
If you assume that you might lose your production site, then you need to consider how you will cater for replacement hardware. Ideally you want your replacement hardware to be similar to the production hardware, as this greatly eases the restore process. You can wait until disaster strikes to buy replacement server hardware, or you can purchase server hardware in advance. If possible, it's advisable to buy the hardware in advance. There are both advantages and disadvantages to this approach:
- Fastest possible recovery time
- Advance knowledge of what hardware you have
- Guaranteed availability
Don't forget that your hardware needs to include not only servers, but also tape drives where necessary.
Once your hardware is sorted out, you need to consider your data. You need to ensure not only that you have all the data you need, but also that you know where it is! (BE report for this). Information on this will appear in a separate article, deciding what to backup (yet to be written)
Obviously if you are considering that you might lose your production facility then you should keep any DR hardware at a separate location, along with copies of your data (link to BE vaulting).
Now that you've got the hardware and software sorted out, you need to make sure that you can make it all work. If you have enough similar hardware, your recovery plans might be simple enough that all you need to do are straightforward restores. If you don't have enough hardware, you may have to consolidate some of your production servers to a single DR server. Whatever you need to do, make sure you document the process thoroughly – it might not be you who performs the process. As part of the process, make sure you have any information you might need, such as passwords and encryption keys. Creating your DR process might take several goes, but keep going until you're confident that it's right.
Once you have your DR process sorted, test it. Then test it again. Make sure that it is updated to reflect changes in the production configuration, and is tested regularly