On 2007/03/28 17:52, rezidue wrote: > When the hosts connect to each other a full prefix table is sent but then > almost immediately it's neighbor starts withdrawing prefixes. On one host I > jump from 210k prefixes in the initial connection to only 59k after all of > the withdrawals. On the other host I go from 210k to 197k which isn't as > bad but I'm still unsure of why it does this.
It's normal. Where I-BGP is concerned, when one router A sees a better route to a prefix from an internal peer B, it will withdraw that prefix because it knows all the other I-BGP routers must know about it (due to the requirement for full-mesh or RR). If that better route from B goes away, A will advertise the prefix again. `better' is defined by, in order: localpref, shortest AS path, origin, MEDs (for multiple routes through the same AS), then on OpenBGP there are some extensions: E-BGP is preferred over I-BGP, weight, optionally route age - then tiebreakers to make sure there's a definite choice e.g. router ID. (Some routers also consider distance to the nexthop by looking at igp metrics, this is done after evaluating MEDs). > ...just use carp for my gateway. take care with this if you're running PF on the routers, pfsync isn't fast enough to keep up if you have asymmetric routing, you may need to use stateless rulesets. (If you're not running PF and just using carp to provide a protected gateway address, this comment doesn't apply).
