As you might notice, we’ve got a new “Twitter Live Feed” on the left sidebar of Network Performance Daily – this is being maintained by the thoroughly awesome Chandra Hosek, who gets to add the moniker “Twitter Wrangler” to her already impressive resume. You can also follow our Twitter at http://twitter.com/NetQoSLiveusing whatever twitter-following app you prefer.
You’ll notice that we’ve put this way down on the sidebar, so you might have to look for it. This is for a very important reason: Twitter’s servers have this annoying habit of occasionally seizing up and drooling on themselves at random intervals.
See, we would have liked to put the Twitter feed prominently on the page, but because many browsers, including Firefox 3, load the page line by line, any content below after the slow loading Twitter feed resolves gets delayed. Placing Twitter at the top of the page would mean the entire site would take seconds to load.
This is actually more than just a specific concern for us – one of the advents of the 21stcentury is the idea that we get added value from our Web applications by combining with other Web applications – mashups.
Many Web 2.0 services, Google maps, data from Craigslist, etc, provide APIs so that third parties can use the applications to develop their own applications or add functionality to their Web apps and Web sites – creating mashups of multiple applications.
So now, instead of having all your apps running in the same place, the apps are running from multiple different places which network engineers have little control over. And in many cases this can result in a delay to the end-user – because your performance is based on the performance of other computers, other networks, other hardware.
Some of the apps that contribute to these mashups may be in turn based on multiple services themselves. The problem can compound itself.
So this is a concern that should be on any network engineer’s mind when talking to the application developers.
No comments yet.