On Wed, Aug 13, 2008 at 9:30 PM, Ed Leafe <[EMAIL PROTECTED]> wrote: > On Aug 13, 2008, at 8:25 PM, Ted Roche wrote: > > I've heard it's more of a software scaling issue, but not being on > the inside, I'm just taking others' word for it. The Rails app that > runs Twitter was never intended to handle the kind of load that it does.
Could be. Hardly any of us anticipate having to scale to this size. Often, you can throw hardware at a problem to get correct code to scale, but that can only take you so far. My understanding was that the initial design had every individual refresh requerying each and every queue that user is subscribed to determine if there was new content to display, rather than taking the opposite tack and let each post update all listening queues. With fast reads and slow writes, and a ratio of lots of refreshes per post, the former can be faster at low volumes (and probably cheaper for a startup), but eventually volume catches up with you. I understand there were some changes in their IT group. I'll bet they're building an interesting architecture queuing posts to subscriptions. I can't see Indenti.ca giving them a run for the money unless someone's willing to throw a lot of money at them. What's the business model for this stuff? -- Ted Roche Ted Roche & Associates, LLC http://www.tedroche.com _______________________________________________ Post Messages to: [email protected] Subscription Maintenance: http://leafe.com/mailman/listinfo/profox OT-free version of this list: http://leafe.com/mailman/listinfo/profoxtech Searchable Archive: http://leafe.com/archives/search/profox This message: http://leafe.com/archives/byMID/profox/[EMAIL PROTECTED] ** All postings, unless explicitly stated otherwise, are the opinions of the author, and do not constitute legal or medical advice. This statement is added to the messages for those lawyers who are too stupid to see the obvious.
