New Year Technologies
This is the time of year when all the pundits start making their “technologies to watch in 2009” predictions. I’m lousy at this game – I bought a Playstation 2 just before the Wii shipped, but on the grounds I can’t be wrong all the time, here’s where I’m placing one of the small number of bets I’m prepared to make in ‘09.
Cloud computing is coming, no points for that prediction, but with cloud computing come some performance challenges that need a technical solution. We’ve had WAN optimization technologies and products for some time, and they try to squeeze more stuff through the pipes. All these products throw a barrage of techniques from caching, to compression, to spoofing, and for many applications the net (pun intended) result is quite impressive. It only gets you so far though, and there’s a category of optimization technology that’s very relevant to workflow. Application Accelerators such as those from F5, Cisco and Replify, go further and do really smart things with the large lumps of content (i.e. documents) that typically get moved around with business processes. Have to edit a document? Ok first time you access it you incur the pain of downloading the whole thing (albeit compressed); but next time you need it, you’ll only get the changes since you last viewed it. Edit it and save it? Well only the changes get shipped back to the server. The net bandwidth and latency reduction is huge for any process involving content. I think we’re going to see new products, and rapid deployment of such products not just for SaaS but for any company with peripatetic workers – and that’s most companies these days.
Ultimately bandwidth always runs out – it’s simply not cost effective to buy enough to ensure that congestion never occurs, and in practice, if content is traversing the “real” intranet, or a remote, possibly cellular, connection, you’re going to get congestion some times. What’s needed is intelligent predictive technology that identifies incipient congestion and uses that layer 3 information to take action at the application layer. Take streaming content such as audio and video for example; the user will always prefer to see a reduction in the quality to a start/stop rebuffering experience, so a proxy that can see the threat of congestion looming and then tune down the streaming bit rate to avoid it, becomes a compelling concept. All the commercially popular streaming servers have means to do this, be it codec switching, bit rate negotiation, frame thinning or something else. I don’t see such products out there today, but I know someone who is building one, and I think it’s worth a bet. Perhaps not too relevant to BPMS today, but I don’t think it will be long before we see business processes which use richer, and streamed, content.
About the Author: Paul Moorhead is the Product Manager at Singularity.
Author : Paul O’Neill
Tags: BPM, Cloud_computing, Processes, SaaS, Workflow