Every ISP who is really serious uses IS-IS because of it's bare-bones implementation (not as complex as OSPF) and its interaction with BGP (Overload bit)
98% of the ISP's I've worked with use OSPF and BGP.
The bare-bones implementation of IS-IS is a drawback for
these "serious", but smaller, ISPs.
Plus ospf, if configured properly(address summarization, redistribution) is easier to configure and manage.
If you have hardware that won't handle ospf when configured properly it won't handle IS-IS either.
In short, I disagree.
A lot of major ISP's run IS-IS instead of OSPF. Maybe not because it's simpler to implement but probably because IS-IS was the first to be implemented.
IS-IS is quite widely used, due to the fact that providers normally run their own SDH transport network, this uses the OSI side of IS-IS for management purposes, whereas the IP backbone uses the IP part of IS-IS
actually Perlman wrote IS-IS after Moy wrote OSPF (by some time) The implementation has more to do with avoiding BGP forwarding before the IGP has converged (Overload Bit) than anything else.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.