All 1 entries tagged Resnet
February 16, 2006
Having only skimmed the update yesterday, today I thought I'd have a good look… Mmmm
Results revealed that delays in response times only occur at times of peak load when the network is congested and we have therefore usefully ruled out the possibility of there being an underlying problem with the handling of DNS.
Which surely reveals an underlying design fault with resnet itself – heck if DNS requests can't get through what chance have my WoW packets got? Seriously…
We are also concerned that the main ResNet firewall is overloaded and are now focussing our attention on this area. As a first step towards reducing the load, we removed the limit on the maximum number of sessions that an individual IP address can set up. This made no difference. We now propose to install an additional firewall to double the capacity.
Translation : we are not sure why resnet is still slow so now we will have a tinker with the firewalls. Two firewalls should be faster than one right?
Seriously though when did the capacity of the firewalls get investigated? Surely firewall logs would be one of the first places to look… oh well.
Aghast with the information so far I almost died when I met the "timetable for change" as I call it:
Thursday 23 February, 9.00am – 10.00am : Implement highest possible priority for the handling of DNS traffic on all components at the core of the network.
Now surely the network is already setup to treat DNS requests as high priority? Maybe not…
I must admit since the very start of the year I have been very vocal to my fellow students about ResNet, but never actually gone along to any of the focus meetings. That will change as I have read only 2 people turned up to the last few…
Ok I admit that finally we are moving somewhere with Resnet but I mean come on, 4 months to get to this point is a frankly a big fat joke which is further compounded by the lack of any proper compensation.
Does anyone have a detailed Resnet network topography to hand?