Two more questions here:
st
Route aggregation is the grouping of smaller routes together into advertisements that are smaller in number (count) then all of the individual routes. Generally speaking, you would use 10.0.0.0/16 as an aggregate to 10.0.1.0, 10.0.2.0, 10.0.3.0, etc, which you may have configured as vlans in a datacenter. Summarization can refer to many things, but can be configured on specific routing protocols to automatically summarize by the Class of the network (based on first octet). Summarization can also be configured manually, either by commands like 'ip summary-route eigrp', or in BGP by creating a route to null0 for the summary and putting in the network statement for the aggregate network or supernet.
To answer your second question, 192.168.2.1 with a 255.0.0.0 subnet mask would mean that router is advertising 192.0.0.0 - 192.255.255.255 via its routing protocol,. Presuming you are asking about a situation where 192.168.2.1 255.255.255.0 is a interface address, and you want to summarize a larger range than this, you would put summaries towards the WAN or Internet (usually summaries configured in the core and then passed into the WAN routers, although some folks prefer to put the summaries on the WAN routers themselves). Using a mask of 255.0.0.0 would most likely be a very bad idea, unless of course you really have a datacenter with 16,000,000 hosts.
At their core, summaries are used to keep the routing table manageable and clean. If you have 10 datacenters, and you were being clean, you could have about 10 or so routes in your table, before WAN links or default routes. If you did not summarize, you would have 10 x each vlan or subnetwork in each datacenter. Particularly when you are dealing with retail companies with hundreds or thousands of stores connected via point to point links, it was easy in the old days to degrade performance on a router or link with all of the updates from the various routes. This is why it is best to use stub, summarization, and default information originate to keep your tables short and succinct.
snipped-for-privacy@hotmail.com schrieb:
To add to Trendkill's very valid answer:
Classes are dead. It's only out of habit that people still frequently use a /24 mask with addresses in the 192.168 range. All networks are classless nowadays in the sense that you cannot reliably deduce the netmask from the first few bits of the address - you need to be told what it is. In particular, networks using addresses from the 10.* and
172.16 to 172.31 private ranges rarely use the /8 or /16 mask respectively that would have been associated with them in the old classful world.HTH T.
Summarization also has benefits in terms of convergence.
Using Trendkill's example, if you had a summary route of 10.0.0.0/16 in your table, you would not be aware of the status of individual "summarized" routes such as 10.0.1.0, 10.0.2.0, 10.0.3.0, which means you would not have to reconverge every time there was a change in status of one of these routes.
A flapping interface on a "summarized" network would not result in you having to recalculate an entire OSPF topology (for example).
It provides stability, and security.
Best regards, News Reader
snipped-for-privacy@hotmail.com wrote:
Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.