Hi Ruben, Sounds exciting, I had it in mind to try something like this too, but a little different. Some questions for you:
1. Network congestion is a big issue for us. Are you ingesting to a central admin service then farming it back out to the capture agents, or are you ingesting directly to the agent the media is on? This latter method could lead to huge performance gains. 2. What hardware are you using for your agents? Did you try the reference one to see how that would work? It's the one I've been considering. 3. How (or are) you scheduling the capture agents so they don't process while they are capturing? Regards, Chris > Hi all, > > I'm just resending this e-mail that was previously sent only to the > "matterhorn" mail list. I hope you find it interesting. > > Best regards > Rubén Pérez > > El 8 de septiembre de 2011 14:21, Rubén Pérez > <[email protected]>escribió: > > > Hi all, > > > > I proudly announce that the University of Vigo is going to start > > using its new "advanced" pilot (this is to distinguish from the > > previous, much smaller-scale pilot). > > > > Our new pilot consists of 10 fully-equipped classrooms and three > > videoconference rooms scattered in different buildings along the > > three university campuses (which are up to 100 km away). The pilot > > will record an average of 100 ~ 200 hours per week and will be in > > test period for the Fall Semester. If, as we expect, the results > > are good, we will finally go into production for the Spring > > Semester, adding new rooms and courses to the existent ones, thus > > additionally extending the number of hour recorded. > > > > We are basing our pilot in the freshly-baked 1.2 Release, in > > combination with a highly-customized version of the capture agent. > > But we are also using a non-standard deployment that has shown very > > promising results in our preliminary tests: the capture agents will > > function as workers in their idle periods, thus creating a rather > > large network of physical workers, instead of a few very powerful > > --perhaps virtual-- worker machines. This experiment implies using > > slightly more powerful machines for the capturers/workers, but we > > believe it's worth because physical machines are more efficient for > > CPU-intensive task than their virtualized counterparts. > > > > It would be very nice if you could share your opinions on this > > approach we are taking, as well as point out the possible drawbacks > > you think it has. Our tests have not found any important issues for > > the time being, but we could use other sources of opinion, > > experience and criticism to improve our pilot as much as we can. > > > > Thanks for your attention! > > > > Best regards > > Rubén > > -- Christopher Brooks, BSc, MSc ARIES Laboratory, University of Saskatchewan Web: http://www.cs.usask.ca/~cab938 Phone: 1.306.966.1442 Mail: Advanced Research in Intelligent Educational Systems Laboratory Department of Computer Science University of Saskatchewan 176 Thorvaldson Building 110 Science Place Saskatoon, SK S7N 5C9 _______________________________________________ Matterhorn-users mailing list [email protected] http://lists.opencastproject.org/mailman/listinfo/matterhorn-users
