Reliability benefits of our own hardware
With our own hardware we can provide things which were not previously possible.
Reliability will be vastly increased, some will be realized immediately upon transfer, some will be realized over time.
The instant benefits are:
- Data redundancy: 99% of data loss cases should be eradicated, all servers will have redundant storage
- HW redundancy: We can simply boot you onto another HW node if available, hastening recovery from CPU/RAM/Mobo/PSU failures.
Over time we can potentially also add Dedicated server upgrades/downgrades, Dedicated server quick storage upgrades and purely a backup service, and full mirroring data backup feature to off-site location.
The last one is tricky and might take more than a year to realize, as the off-site location needs to be connected as well with at least 10Gbit, even with 1Gbit the vast amounts of data we need to transfer would take way too long to be relevant. 1Gbit is able to transfer just 300-310Tb a month, so it would likely take more than a month for the initial sync for the first servers, never mind when we reach full production capability...
We will likely also proceed to offer storage orientated services, such as backup service as a VPS for professional usage, likely starting from just 10€ a month with 1Tb storage. 1/10th the price of Rsync.net.
In this service there would be no restrictions how many computers, what kind of data etc. unlike with the "unlimited" ones which limit you to 1 computer, and excludes large, esp. incompressible data files such as VMWare images, large video files etc.
There really is no service like this currently in offering, and we would have gladly used something like that in the past to provide data redundancy at a sensible price range. Of course scaling up to 30Tb, later on upto petabytes.
Wednesday, May 22, 2013
