network guys - HSRP

[ScottJ]

So, with passive Eth circuits (cross-connects) HSRP works fine switching from primary to backup path when primary path fails, but our circuits here are carrier metro-GE and carrier does not support link-pass-through / fault-propagation, so if down on my side it can be up on customer side.

So far, only when both client and provider primary port are link-down, HSRP backup path starts working.

I didn't think HSRP path-switch required link-down both client and provider (on primary interfaces both ends) because of HSRP Hello's going between the 2 routers (via customer L2 switch)...but the carrier cannot support link pass through, so this sucks. I need to get the HSRP redundancy function working with these metro-E carrier circuits, there is no alternative carrier path.

I'm reading up HSRP http://www.faqs.org/rfcs/rfc2281.html

anyone have successful HSRP redundancy-path-switching working on segmented ethernet WAN between 2 provider routers and the customer switch?

it looks like this:

r1(cisco) r2(cisco)
gi-xxx gi-xxx
|(primary) |(backup)
[ carrier site-a ]
| 2x Metro-GE |
[ carrier site-b]
\ /
\ /
Cust HP L2 switch
| | | |
| | | |
Cust servers


thx

[ceyko]

I'm having problems following your situation here a little. I get the basic idea so I'll throw out a few things but without a diagram I'm not going to "get it"

1. HSRP track command

http://www.cisco.com/en/US/docs/routers/asr9000/software/asr9k_r4.2/addr_serv/command/reference/b_ipaddr_cr42asr9k_chapter_0110.pdf

2. In some of those situation I use PBR with IP SLA monitoring...have not watched these videos but google PBR with IP SLA or ...

http://www.youtube.com/playlist?list=PLA2431A989529805C

basically you can monitor IP addresses and if they lose connectivity it'll use backup routes...etc...etc. Some of this is dependent on hardware, licensing and architecture.

[ScottJ]

Thx for looking!

I'm looking at hsrp track now actually, maybe it will help.

Basically, I can't get HSRP traffic to fail-over right because of the link-up condition no matter what on both my router interfaces, like when the customer at remote site takes one of their uplinks down. It's because of the separate ethernet segments on the WAN (no link fault propagation happening)

I'm having problems following your situation here a little. I get the basic idea so I'll throw out a few things but without a diagram I'm not going to "get it"

1. HSRP track command

http://www.cisco.com/en/US/docs/routers/asr9000/software/asr9k_r4.2/addr_serv/command/reference/b_ipaddr_cr42asr9k_chapter_0110.pdf

2. In some of those situation I use PBR with IP SLA monitoring...have not watched these videos but google PBR with IP SLA or ...

http://www.youtube.com/playlist?list=PLA2431A989529805C

basically you can monitor IP addresses and if they lose connectivity it'll use backup routes...etc...etc. Some of this is dependent on hardware, licensing and architecture.

[Grant]

Who's the circuit providers?

[mustang_revival]

Yes your quickest path would be setup a ping health check to failover when your primary leg can't pass traffic.

Other than that utilize a internal BGP active active setup which would fail-over when the BGP neighbor was down.

[ScottJ]

we're looking at ping health, not sure I should disclose the ckt provider but they are metro-ethernet GE's, that only support untagged or tagged L2 traffic.