Re: ospf fib reload and rtm_desync

2011-08-01 Thread Stuart Henderson
On 2011/07/29 14:33, Stuart Henderson wrote: > Here's a diff. Rather than reloading directly on receipt of > RTM_DESYNC it arms an evtimer so that it defers reloading until > 1sec after the last desync message. (Randomly chosen value and > might want to be larger, but it serves well as a demonstrat

ospf fib reload and rtm_desync

2011-07-29 Thread Stuart Henderson
Fairly long mail with loads of boring debug output, skip this if you're not interested in ospf / route table dumps. Here's a typical ospfd startup on a machine with a full route table already present. Note it does 33 full table dumps. It's on my home hack network, just a couple of ospf speakers, t