MTO and RTO in a Cloud Backup Context

A couple of interesting cases came up recently about differences in cloud backup services, and the effect that this might have on MTO and RTO. As a reminder, maximum tolerable outage (MTO) is the maximum time you or your organisation can afford to be without a given system or resource; recovery time objective (RTO) is your goal for getting a system back in operation. The relationship between the two is MTO = RTO + work recovery (tackling the backlog due to the outage, for instance). In an IT context, whether or not you are over or under the limits set for these measures may depend on how fast you can recover backed-up data – but also on how fast you can back up that data in the first place.

The idea of cloud backup is simple enough. You send your data over the Internet for it to be captured, organised and stored by a cloud services company. When you need it, you get it back the same way. So far, so good – except that some inexpensive services run out of steam after a certain volume of data has been transferred, with data transfer rates slowing significantly thereafter. If you can’t back it up quickly enough, you won’t have it available to pull back in times of trouble: hence the potential risk in relation to MTO and RTO.

Some providers apply “data throttling” to slow down transfer of data over the network. An example is backing up a 500 Gigabyte hard disk (common enough in PCs nowadays) on a service that applies data throttling when the volume transferred reaches 200 Gigabytes. Backup times on one service stretched out for as long as nine months. So the moral of the story is probably that once you have determined what you need in terms of MTO and RTO, then choose a level of cloud service that is appropriate, rather than just the cheapest one.