Question for the networking gurus out there. What options do you have to provide redundancy for the failure of a network switch, especially, if the switch doesn't actually go off-line but stays on-line but acting all crazy?
We have some mission-critical servers shared between multiple campuses co-lo at a level 4 data center, and pay a ton for this co-lo. Anyway, a few hundred senior people have been sitting around contemplating their navels all day, because the apps hosted on these servers are inaccessible.
It appears that the problem is that one of the network switches that was co-located went crazy, and started dropping about 95% of packets. Enough to be unusable, but not enough to cause any links to get failed out.
It needed someone from networks to physically drive down there with a replacement switch.
We have some mission-critical servers shared between multiple campuses co-lo at a level 4 data center, and pay a ton for this co-lo. Anyway, a few hundred senior people have been sitting around contemplating their navels all day, because the apps hosted on these servers are inaccessible.
It appears that the problem is that one of the network switches that was co-located went crazy, and started dropping about 95% of packets. Enough to be unusable, but not enough to cause any links to get failed out.
It needed someone from networks to physically drive down there with a replacement switch.
