question about redondant supervisor configuration

christianpho · ‎02-17-2008

Hi

I'm currently preparing my BCMSN certification exam. Follow my reading I need more explications on some topics. Also I'm am a French people and some time because all documentation I have are in English some times it is the English it self my understanding problem.

1# When I use redondant solution for supervisor engine in multilayer switch like a 6500 serie I have probebly to use identical unit in active mode and standby more with same IOS installed on them to make sure all I'm using when everything working I will able to use this configuration when I will use the redundant unit. (is it work work is the redundant unit is different then the active unit, with reduce capacityâ€¦.)

Is it exist some reason where I don't want use the better redundant mode available for the model I'm currently working with ?

2#Here copy paste text from my CCNP certification guide :

" TIP Sometimes the redundancy mode terminology can be confusing. In addition to the RPR,

RPR+, and SSO terms, you might see Single Router Mode (SRM) and Dual Router Mode

(DRM).

SRM simply means that two route processors (integrated into the supervisors) are being used,

but only one of them is active at any time. In DRM, two route processors are active at all times.

HSRP usually is used to provide redundancy in DRM.

Although RPR and RPR+ have only one active supervisor, the route processor portion is not

initialized on the standby unit. Therefore, SRM is not compatible with RPR or RPR+.

SRM is inherent with SSO, which brings up the standby route processor. You usually will find

the two redundancy terms together, as â€œSRM with SSO.â€

"

This let me confuse the DRM mode mean 2 processors are running in the same times like I haved in SSO where the redundant unit is fully boot and this mode is the mode where the redundant unit is the more ready to take all the charge of work when the active unit failed. Following the first part of this text I understand completly the inverse of what is was explain in the second part of the same text :

Here what is my understanding :

"Although RPR and RPR+ have only one active supervisor, the route processor portion is not

initialized on the standby unit. Therefore, DRM is not compatible with RPR or RPR+.

DRM is inherent with SSO, which brings up the standby route processor. You usually will find

the two redundancy terms together, as â€œDRM with SSO.â€"

Someone can help me to explain what I missed understand ??

Thans a lot !

Joseph W. Doherty · ‎02-17-2008

Looking at some documentation, http://www.cisco.com/en/US/prod/collateral/switches/ps5718/ps708/prod_white_paper09186a0080088874.html, it appears DRM was an earlier approach, not available on current gen sups (sup32/sup720), to provide fast failover. Effectively, each sup was running as an independent router watching each other via internal HSRP.

SRM modes provided slower failover until the later NSF/SSO.

paul.matthews · ‎02-18-2008

As long as hardware and software support it SSO is almost certainly the way to go.

The two supervisors will share the FIB information, and on failover the standby will have alled cached packet forwarding entries available, meaning that as a test you could set up two PCs running a ping and possibly not lose a packet.

If the switches are running L3 it gets a little more interesting. You then need to look up NSF awareness and configure the neighbours to be aware of the SSO/NSF capability - basically you add a delay between losing the OSPF (or whatever) neighbour and actually actioning the loss.

With a static route in place to cover the few seconds where the new supervisor comes on line and the routing protocol converges.

I am demonstrating that shortly with a live video stream...