rs client and bgp soft reload out
Hi guys, I have 2 BGP related questions that would like to confirm with you guys: 1. Can we run the route server without setting "rs client" BGP configuration in each client? I can bring up the eBGP as well without that setting in place. 2. Birdc reload out "protocol_name" command will do soft reload on the exported routes to that particular neighbour. However this will reset the uptime. Any idea whether we can keep the uptime? It is very worrying while doing the troubleshooting if the uptime is not reflecting the correct uptime. Thanks, Jimmy
On Thu, Jan 24, 2013 at 11:55:39PM +0800, Jimmy Halim wrote:
Hi guys,
I have 2 BGP related questions that would like to confirm with you guys:
1. Can we run the route server without setting "rs client" BGP configuration in each client? I can bring up the eBGP as well without that setting in place.
Hello You could, but in that case your BGP ASN would be in AS path, which is usually considered unwanted for route servers. Also MULTI_EXIT_DISC is reset when forwarded as usual, but kept when forwarded as route server. And there is a different default for IPv6 and missing link-local address (see 'missing lladdr' option).
2. Birdc reload out "protocol_name" command will do soft reload on the exported routes to that particular neighbour. However this will reset the uptime. Any idea whether we can keep the uptime? It is very worrying while doing the troubleshooting if the uptime is not reflecting the correct
You mean 'state since' in 'show protocols'? This is not exactly uptime of the BGP session, but the uptime of current protocol state. which changed from 'up' to 'feed' and then back to 'up' during 'reload out' (while the BGP session is up for the whole time). We could probaly add some other counter to track BGP session uptime. -- Elen sila lumenn' omentielvo Ondrej 'SanTiago' Zajicek (email: santiago@crfreenet.org) OpenPGP encrypted e-mails preferred (KeyID 0x11DEADC3, wwwkeys.pgp.net) "To err is human -- to blame it on a computer is even more so."
Thanks for your input Santiago :p For the 2nd question, yes I am referring to 'state since' in show protocols. So that is the uptime of current protocol state. It will be great if we can keep track the BGP session uptime since it will help a lot during troubleshooting.. Cheers, Jimmy On 25/1/13 4:11 AM, "Ondrej Zajicek" <santiago@crfreenet.org> wrote:
On Thu, Jan 24, 2013 at 11:55:39PM +0800, Jimmy Halim wrote:
Hi guys,
I have 2 BGP related questions that would like to confirm with you guys:
1. Can we run the route server without setting "rs client" BGP configuration in each client? I can bring up the eBGP as well without that setting in place.
Hello
You could, but in that case your BGP ASN would be in AS path, which is usually considered unwanted for route servers. Also MULTI_EXIT_DISC is reset when forwarded as usual, but kept when forwarded as route server. And there is a different default for IPv6 and missing link-local address (see 'missing lladdr' option).
2. Birdc reload out "protocol_name" command will do soft reload on the exported routes to that particular neighbour. However this will reset the uptime. Any idea whether we can keep the uptime? It is very worrying while doing the troubleshooting if the uptime is not reflecting the correct
You mean 'state since' in 'show protocols'? This is not exactly uptime of the BGP session, but the uptime of current protocol state. which changed from 'up' to 'feed' and then back to 'up' during 'reload out' (while the BGP session is up for the whole time).
We could probaly add some other counter to track BGP session uptime.
-- Elen sila lumenn' omentielvo
Ondrej 'SanTiago' Zajicek (email: santiago@crfreenet.org) OpenPGP encrypted e-mails preferred (KeyID 0x11DEADC3, wwwkeys.pgp.net) "To err is human -- to blame it on a computer is even more so."
participants (2)
-
Jimmy Halim -
Ondrej Zajicek