Tuesday 9 January 2007

Business Continuity and Web-based applications: Sync and Distribute

A few days ago I had a wibble moment with some of the Web-based tools use on a regular basis - for a short period (minutes if that) I couldn't reach Gmail or Google search. Not sure if the problem was the Google end, with my ISP or somewhere in between. Anyway, with recent Internet outages in Asia, it again made me ponder the business continuity issues of Web-based applications.

I'm not the only one to experience wibbles with Web-application services recently, but I think Charlie Wood is (now) approaching the problem with the right attitude - he describes his proactive approach to business continuity:

"I keep a copy of my most important data from salesforce.com on my laptop using Spanning Salesforce, and thanks to the magic of RSS, it's always up to date. Similarly, I synchronize all of my Google Calendar events with iCal running on my laptop using Spanning Sync. That way if Google Calendar is offline I still have full access to my calendar."

However, I think synchronise might be the wrong activity to emphasis - maybe what we should be trying to do is distribute data and application further to ensure business continuity:

  • Use multiple data stores in different physical locations and on different physical devices; and
  • Use multiple applications and service providers to access these data stores.

Synchronisation is of course important for keeping the data accessed current, but I can't help thinking of JP Rangaswami's experience of his blog's Web hosting failure and how his content was recovered again:

"The community has been fantastic, coming up with rich and varied suggestions as to how I could salvage the blog. A number of you scraped Google caches and sent the salvage on, particularly Chris and Doc. One, Myrto, had a complete set of my posts in Outlook via Newsgator. Some of you, particularly Malc, pointed me at the feedburner cache where the last 73 posts were available. My Mac account had a faithful copy of all comments received for moderation. While Google, Wayback Machine and Feedburner were less than complete, Niall found that Alexa had the complete store."

No comments:

Post a Comment

Note: only a member of this blog may post a comment.