X-Git-Url: https://git.sommitrealweird.co.uk/quagga-debian.git/blobdiff_plain/191fe7a34582876de01d3e62c2a6587baf59a283..064d9c633233495319bcaa66335ea3d24c0bd7a1:/pimd/WHY_SSM?ds=inline diff --git a/pimd/WHY_SSM b/pimd/WHY_SSM deleted file mode 100644 index 2e8c966..0000000 --- a/pimd/WHY_SSM +++ /dev/null @@ -1,32 +0,0 @@ -WHY SSM - -Benefis of PIM SSM over PIM SM ------------------------------- - -- SSM consumes minimum link bandwidth -- SSM simplifies multicast address management (specially important for - inter-domain multicast) - - SSM (S,G) channels easily provide unique per-application addressing - - SSM does not require MSDP between PIM domains -- SSM does not suffer instabilities from traffic-driven SPT switchover -- SSM is not suscetible to DoS attack from unwanted sources -- SSM does not use RP. Some RP issues: - - RP is possible point of failure - - RP demands redundancy management - - RP may require PIM dense mode support for RP election - - RP is possible performance bottleneck - - RP may demand lots of extra management -- SSM can be deployed in an existing PIM SM network (only the last hop - routers need to support IGMPv3) -- SSM is easier to deploy and maintain - -PIM-SSM drawbacks ------------------ - -- SSM requires IGMPv3 support on both receivers and last-hop routers -- SSM may be memory intensive when managing (S,G) states for - many-to-many multicast distribution -- SSM will keep (S,G) state as long as there are subscriptions from - receivers, even if the source is not actually sending traffic - ---EOF--