You are browsing a read-only backup copy of Wikitech. The live site can be found at wikitech.wikimedia.org
at the moment, we have two core switches in tampa, csw1-pmtpa and csw4-pmtpa. we would like these to be redundant, such that if one goes down, the other can take over. the first part of this is to configure two routes to our ISP.
the layout looks like this:
hosts hosts | | | | csw1------csw4 \ / \ / \ / powermedium
the two switches peer with powermedium using BGP. if both are up, half the traffic goes through each. if one is down, all the traffic goes through the other one. additionally, the switches peer with each other. this means that if one switch is up, but its fibre is down, it knows how to send internet traffic to the other switch.
there is no host redundancy yet (this is waiting until the ISP part is finished).
so far we have:
- connected csw1 and csw4
- set up two fibres to PM, one from each switch
- set up BGP on csw4
we are waiting to:
- set up BGP on csw1
- set up peering between both switches