Bandwidth management followup
Bandwidth management teething problems (previous) :: New permanent job in the ResNet team (next)
The problems we were experiencing with our new bandwidth management appliance (see bandwidth-management-teething-problems) seem to have highlighted a design flaw in the appliance itself. We're working with the vendor to fix the problem, although we're not expecting a patch from them for another two weeks or so.
Latency and packet loss are now back to the same levels that they were in January. However, these levels are still much higher than they were at this time last year. This is mainly because there's a *lot* more traffic on ResNet than there was this time last year!
We will be making some changes next week which we hope will improve the situation further, we will be documenting details of the changes (Most importantly the details of the times when we put them in place) on the blog.
We've set up a new category for bandwidth related posts, so if latency/packet loss etc is of special interest to you (eg if you're a gamer) you might want to keep an eye on the bandwidth management category.