Chuck Hollis pondered whether using a remote DR facility could ever be consider ‘Cloud Bursting’ and much conversation is ensuing along the lines that ‘Cloud Bursting’ is a marketing thing which currently doesn’t exist and won’t exist until applications can be architected which automagically scale and move themselves to utilise capacity where-ever else it maybe. I am paraphrasing some much brighter people than me who know a lot more about Cloud than a mere Storagebod but that’s kind of the message I took away.
Anyway, what Chuck was pondering is not exactly new; for decades, we have moved workloads about; sometimes moving them temporarily to a DR site or second site to free up capacity for a transient capacity demand or whilst waiting for a capacity upgrade. Mainframe houses are/were pretty well versed in this; shifting workloads at peak times and then bringing them back when the crisis has averted.
It’s not really bursting; bursting is something which just happens and is dynamic, immediate and exciting. This sort of workload management is somewhat akin to bailing out a boat or perhaps transferring a liquid from a now too small container whilst you either stem the flow or get a bigger container.
And yes, you could use the public Cloud as your temporary container but you could also use your DR site, perhaps your development kit…but to ‘Cloud-Wash’ it as ‘Cloud-Bursting’ is probably pushing things a little far. So perhaps ‘Cloud-Bailing’; simply on the grounds that it sort of brings to mind of something which is neither elegant and a little haphazard.
Or perhaps we could just call it Workload Management and consider that its a general discipline which could be applied equally to Cloud and to more traditional IT?
Being a support guy I’m forced to say: Don’t use DR sites for anything other than DR. …but this could be superstition, too… :o)