Maximum number of configurable RADIUS servers on Cisco IOS. Anyone knows?

A friend of mine raised an interesting question.

What is the maximum number of configurable RADIUS servers on Cisco IOS platform?

I failed to get this exact number after dozens of Google searching. Can any one help me about this? Thank you!

 

By the way, for Cisco NX-OS, the maximum number is “64”, which is described in documents like this one.

http://www.cisco.com/en/US/docs/switches/datacenter/sw/4_1/nx-os/security/configuration/guide/sec_radius.html

Client PC's Redundancy: Wireless LAN, also backup Internet's traffic

His happy trip ending, and so was mine (DSC_0237),
originally uploaded by Li-Ji.

Continued from previous post. Have you come up with the answer?

The key is still the "more specific route": we only have to craft some routes more specific than the "default route"! The default route itself is indeed the least specific! Any routes with just one bit more specific, is enough to be more specific!

---

OK! Let me give you the answer: we only have to add two routes 0.0.0.0/1 和 128.0.0.0/1, pointing both of them to Wired LAN's gateway IP address, done! Now the Internet traffic will go through Wired LAN in normal time. It goes through the Wireless LAN only when Wired LAN is broken!

Why these 2 routes? We can clearly see it by looking at its binary form:

33          2          1          0       33          2          1          0
1098765.43210987.65432109.876543210       1098765.43210987.65432109.876543210

00000000.00000000.00000000.00000000, mask 10000000.00000000.00000000.00000000

10000000.00000000.00000000.00000000, mask 10000000.00000000.00000000.00000000

Of course, we can also come up with 4 routes, or even 8 routes to play the same trick! Two is the simplest one!

Interesting, isn't it! This trick is quite useful in many cases! We might see it again very often!

Client PC's Redundancy: Wireless LAN

Yellow Flowers (DSCN0947),
originally uploaded by Li-Ji.

To provide Client PCs with network backup redundancy, Wireless LAN has a unique advantage that all Wired LAN cannot provide at all: Automatic Recovery! Windows will do this by itself to find another available "Access Point", automatically! That's why I emphasize Automatic Recovery: user will not even notice the network has been down for a moment because of single network device's failure, when all the client PCs use Wireless LAN only to connect to the network!

However, Wireless LAN today still has it technical limitation of "Bandwidth". This problem would become more serious if client PCs are located densely close to each other, and each client PC runs applications consuming lots of bandwidth. Using Wireless LAN only on client PC would now be a nightmare for network administrators, if the client PCs should be arranged in this way!

I have another suggestion for you. We can use Wireless LAN as a backup redundancy to Wired LAN, so we can take both technology's benefit at the same time. In the normal scenario the whole traffic goes through Wired LAN. Once the Switch fails, client PCs divert traffic to Wireless LAN for backup. Although at this backup time the network would still be slow because of Wireless LAN's limitation, at lease we can still keep full connectivity to all client PCs.

Can the switchover be automatic, too? Sure! The trick is to make use of "more specific routes"! Assume the Intranet is within the "192.168.0.0/16" range. We first separate both Wired LAN and Wireless LAN into different network of addressing. Then we point the route to "192.168.0.0/255.255.0.0" on Windows to the gateway IP address of Wired LAN, and point the "default route" to the gateway IP address on Wireless LAN side.

You see it! In normal time traffic would go through more specific route through Wired LAN. Only when Wired LAN is down would traffic go through Wireless LAN. Automatically done!

As how to add those two routes in client PCs, we can make use of Window's AD Group Policy. Or even simpler, we just use DHCP to insert the two routes into client PCs.

I believe some of you might come up with another question: this solution only solves the "Intranet" automatic back redundancy. How about the traffic to the Internet?

The trick is very similar. Let me take a breath and you keep watching my blog! I will tell you in my next post!

Redundancy: a great lesson to me!

I believe many you have noticed about some .PKT files and some .JPG files cannot be downloaded or viewed since one day ago. In fact, a serious down time is happening to me my ONLY web hosting provider.

Their attendant of customer service of this provider said they are doing some emergency recovery, and are doing their best, but no ETA (I guess it is Estimated Time of Available)! As the time of this post, the service is still not recover yet! It is already over 24 hours and breaking their 99.9% up time promise!

I have to say sorry for your patience! Because I did not backup some of my important files, some contents would take longer to recover!

I always preached to my students about redundancy in network design . However, I did not do with my contents under this discipline. This time I am having a great (and painful) lessen to me!