On Sat, 2007-12-01 at 07:11 +0800, Ed Greshko wrote: > Mark Haney wrote: > > Ed Greshko wrote: > >> Mark Haney wrote: > >>> Ed Greshko wrote: > > > >> > >> OK....so this is OT. > >> > >> But, could you try rephrasing the question such that it is generic? I > >> still > >> don't have the foggiest idea of what you are asking. > >> > >> Yes, it is late in the day here in Taiwan....but I am a fan of the KISS > >> principle. What do you want to know? > >> > > > > I'm really looking for either a tool, package, other that will track > > video views, bandwidth usage, etc for a customer who is starting get > > into 'viral marketing' (I hate that term, but there you have it) and > > want to track video views like page views. Does that make more sense? > > Yes, that does make much more sense. Sorry, I can be dense at times. > > Off the top of my head, I can't think of any tools for that usage. I'll > mull it over a bit and if I run into something I'll let you know. > > > I was really hoping for general input to that since googling came up > > with not a whole lot. I founded one of the largest streaming companies around. This is an issue that's always been a problem. There's no simple way to track "views" of a stream other than parsing the log files from the server and tracking them that way. You must look for (at least) start events, stop events, pause events, and resume events with the same session ID. You can compute the view time from those. It ain't easy. We process literally millions of log events every day. As far as bandwidth, you can use the log entries (doesn't generally cover the overhead--only the active streams), RRD/Cacti, or if you use Cisco gear, Netflow (you'll have to write programs to analyze that). We use all of those and then some, but we're a commercial outfit. ---------------------------------------------------------------------- - Rick Stevens, Principal Engineer rstevens@xxxxxxxxxxxx - - CDN Systems, Internap, Inc. http://www.internap.com - - - - LOOK OUT!!! BEHIND YOU!!! - ----------------------------------------------------------------------
