Often we all don't get true clarity on the importance of certain issues until you are living them. That is why for instance Dutchess County BOCES attempts to do a yearly, live, full DR test. It is amazing what you find out when you actually force yourself to go production on plan B. Documentation is missing, people have changed roles, the software has been updated. It is truly the example of continuous improvement to be prepared for the real disaster.
One other area that sometimes is hard for business officials to comprehend is the importance of paying for the appropriate level of a service contract for core equipment. I remember Bob having a discussion with a business official who thought 4-hour, on-site service was too expensive and irrelevant. Bob asked the business official their acceptable level of time to recovery. He was told a week. A week means the much cheaper, next day business service was acceptable. Remember that next day business is no weekends or holidays and calls received after 5 pm means they automatically wait another day.
As fate would have it this district had a major outage - on a Friday going into a holiday weekend. The appropriate vendor calls were made. Due to the nature of the failure, the call went in after 5 pm. Since bad things generally happen all at once, Monday was also a legal holiday. That meant that we wouldn't get what we needed to recover the network to start our recovery work until the following Wednesday! The business official started getting concerned as to why this was taking so long. The reason was the lack of a 4-hour on-site service contract.
We got the part on Wednesday and quickly restored functionality. The business official then rather promptly upgraded that equipment to 4-hour on-site warranty!
In evaluating what level of service you require the real question is whether the broken equipment has redundancy or not and whether the equipment being broken puts the district at any undue risk. If you have 3 physical domain controllers, chances are that you can live without one for a while with only minor discomfort. In that scenario next day business warranty probably makes sense. If you have one SAN and it runs the entire district and it has no redundant, replicated SAN, you want that on 4-hour on-site because if that is dead, the district is electronically dead.
If you need help sorting this out, call us and we'll help you figure it all out.
You must be logged in to post a comment.