Management of Ethernet Based Circuits

ritsonr · ‎08-24-2010

Hi... I'm looking for some opinion on this topic, i've done a bit of reading but can't find much about how SPs and Enterprise manage Ethernet based services.

So in a senario where we have:

Site A CPE >> Carrier NTU >> L2WAN >> Carrier NTU >> Site B CPE

If there is a fault condition in the L2 WAN it would be most likly that CPE interfaces would remain in an UP/UP state. If there were resilient links between Site A and B you would not get a Site Down Type Alarm. Obviously any L3 routing protocol would go down but there must be a clever way round paying for a licence to monitor BGP states and one of the ways i have read about which is using tunnel keep alives just seems messy.

Has anyone done anything smart to try and work arround this problem or do we all just buy the bolt on's for our chosen network management platform?

Thanks,

Richard.

paolo bevilacqua · ‎08-24-2010

Actually there is no license to pay. Either routing protocols, IP SLA, EEM, ULD, all these features are included in the

base image 7 license for all platforms. Different customers use different one according to their need and their liking.

ritsonr · ‎08-24-2010

Hi, thanks for the reply... With the network management platform we use, monitoring of BGP events is an extra module which you pay for per instance...

Not sure what you mean by base immage 7 license?

I started looking at IP SLA for this, will have a look at EEM now, not sure what ULD is?

paolo bevilacqua · ‎08-24-2010

Uni Directional Link Detection, another IOS feature for switches.

If you are looking at this purely from netman perspective, post in the dedicated forum.

gatlin007 · ‎08-24-2010

When an OSPF/EIGRP adjacency is torn down or a BGP neighbor goes down the router will create a syslog trap with the proper level of logging.

Send this trap to an open source network monitoring software such as NAGIOS. Configure NAGIOS to filter them and send an email alert based routing events; or toggle a dashboard indicator for the NOC. We should always know the state of our routing protocol relationships regardless of the underlying media purchased from a service provider.

Chris