This has absolutely no relevance? If all the server is doing is answering a call home then it doesn't need much in the way of resources. Granted there should be a backup, but you don't actually recommend anything about how to implement an effective one, your just showing off about what is in your rack.
I realise this is probably their web server as well but still, a website and answering 'calls to home' is not a resource hog. So you could justify running these services on a rented dedicated server from a BIG, mainly trusted host.
I personally would have 2 servers exactly the same. If the 'Active' falls over I would connect to the 'Backup' via KVMoIP and assign it the IP address of the 'Active' server, then I would put the 'Active' server only on an internal VLAN (To exclude external connectivity and to prevent any routing issues with 2 servers having the same IP) and diagnose/fix the problem then remove the IP altogether. It's not automatic, but it allows you to get the site/services backup reasonably quickly from when you are notified they are down (I don't really trust any automated failover systems). However, I have just doubled their monthly running costs.
SO you do always have to remember the business model of TweetAttacks. It is based on a one time fee, NOT MONTHLY PAYMENTS. The cost of hosting is always going to be on a monthly basis so they are potentially always at a loss there. Then remember that this is a business for them, so to make maximum profit they WILL get their costs as low as possible - that's business.
Granted it's not a great situation, but the TA guys HAVE reacted and no doubt have been running around tearing their hair out trying to get this sorted and they seemingly have implemented a solution. They had the right to turn around and say "It's Christmas, We will not answer support request until xyz" but they didn't... I don't even own TA, but you HAVE to give them credit for sacrificing their own holidays.
PS lucky31337 - Floors are raised in DC's so they can run power and network cables. Not to protect against flooding. Granted Savvis may have flood protection systems, but I'm sure it won't be the raised flooring. (Quick cheeky edit: They are also integral for the aircon to work - all the cold air goes under the floors to the cold aisles, the cold air gets sucked through the front of the rack/servers and is then pushed out to the hot aisle, where the hot air rises and is extracted out of from the Data Floor)
And to GiftGuru - Co-locating a server may not be the best option. With a rented Dedicated server any good host will replace any failed hardware within a matter of hours (Usually less than 1). With a co-located server, if hardware fails, it will sit in the rack broken until you either come and replace the component or you send a replacement in and get tech support to replace it.
Last edited:
