[quagga-dev 9636] Re: [PATCH RFC] OSPF vertices memory exhaustion
paul at jakma.org
Mon Aug 6 12:59:13 BST 2012
On Sat, 4 Aug 2012, Joakim Tjernlund wrote:
> This footnote is attached to the backlink check:
> Note that the presence of any link back to V is sufficient; it
> need not be the matching half of the link under consideration from V
> to W. This is enough to ensure that, before data traffic flows
> between a pair of neighboring routers, their link state databases
> will be synchronized.
> So ANY link will do at this stage.
Yes, that footnote is about the SPF exploration stage. There any link back
is fine, and is what ospf_spf_next checks for.
The problem that was being solved was at a later stage.
Nexthop-calculation was getting led astray by a non-bidirectional edge and
terminating, when there were other paths (either direct links back, or
higher cost through intermediaries) that should have done. That was
leading to unnecessary blackholes during convergence.
Paul Jakma paul at jakma.org @pjakma Key ID: 64A2FF6A
The important thing is not to stop questioning.
More information about the Quagga-dev