ASA 5520 Memory Upgrade

For the benefit of those that might be trying to do it ‘on the cheap’, the official part number ASA5520-MEM-2GB= from Cisco ships with memory from Smart, part number SG5722885D8D0FO and is listed here on Smarts Web Site as being :-

UDIMM 1GB 128Mb x72 2.5V 184-pin 400MT/s PC3200 CL = 3  DDR SDRAM

Hope this is helpful to others looking to pay a little less than the £300+ charge for the official part.

** A quick note on this, installing the 2 x 1GB DIMMs will not result in a running ASA with 2.5GB available RAM. It sees it on bootup but ignores the extra 512MB once it’s running so you may aswell replace the 2 x 512MB sticks with the 2 x 1GB sticks as it gives you no benefit.
To top that, in my particular situation, my ASA went into a boot loop and didn’t start the new 8.3.1 code. To fix that I had to break out of the boot sequence into rommon and type boot asa821-k8.bin which was the image I’d upgraded from – so DON’T delete your old image before you’re sure your ASA is going to boot off it properly of you’ll have to do a tftpdnld by setting an interface with IP details and dragging an image from a TFTP server. Pain – in – the – ass. **

Be Cool

Quick Note:-

Since I’ve a second ASA to install/upgrade and put into Active/Standby, I’ve just re-visited this upgrade and found the following post on
Here it states  :- * Note:  The maximum memory supported for the ASA-5520 and ASA-5540 is 2 Gb.  If you install 4 Gb of memory in these units, they will go into a boot loop.

This I think is exactly the problem I had with my 5520, though it only had 2.5GB in it – the original 512MB plus the 2GB upgrade, so I’ll be removing the 512MB from the existing one and the new ASA when I go to do maintenance on them next week.


Know EIGRP Metric – FAST

(I’ve revised this document in June 2012 to be more concise, apologies for any confusion, there were a couple of ambiguous sections in here for sure!)

Right, it’s taken me an hour but I feel the need to share with you and note for my future reference the way EIGRP metric is calculated by DEFAULT.

You can’t really beat Cisco’s whitepaper on the subject of EIGRP once you’ve cleared it up in your head, but I did take a moment to get my shit together and realise exactly what was going on.

Okay, lets ignore topologies as it’s not going to help right here but you are already running EIGRP right? You wanna know how that phat number in the sh ip route or sh ip eigrp topology arrived there, yes?

Okay – check this EIGRP topology

BB1#sh ip eigrp topology
IP-EIGRP Topology Table for AS(1)/ID(
Codes: P – Passive, A – Active, U – Update, Q – Query, R – Reply,
r – reply Status, s – sia Status
P, 2 successors, FD is 2681856
via (2681856/2169856), Serial0/0.2
via (2681856/2169856), Serial0/0.3

The numbers here you’re interested in are the ones highlighted in bold. The first is your Feasible Distance the metric your router your on has calculated to get to the network in question, the second number is the Advertised Distance which is simply this same calculation of metric performed by the router described here as via serial 0/0.3 to reach the network

Okay. For each entry in your topology table that you wish to understand the calculation, the king command here is

BB1#sh ip eigrp topology
IP-EIGRP (AS 1): Topology entry for
State is Passive, Query origin flag is 1, 2 Successor(s), FD is 2681856
Routing Descriptor Blocks: (Serial0/0.2), from, Send flag is 0x0
Composite metric is (2681856/2169856), Route is Internal
Vector metric:
Minimum bandwidth is 1544 Kbit
Total delay is 40000 microseconds
Reliability is 255/255
Load is 1/255
Minimum MTU is 1500
Hop count is 1

The highlighted bold values are the interesting ones BY DEFAULT. The point being hardly anyone shags with the other K values in EIGRP, they are always 0, so 0 x anything is 0. K1 (bandwidth) and K3 (delay) are the only ones that have values and they are both 1.

The trick here is to understand what to do with these two numbers to get your resulting metric.

The bandwidth – which is the lowest bandwidth on the path to the network of interest has to be calculated as such:-
10,000,000 (or as some people say 10 to the power of 7) divided by bandwidth (as it’s displayed in the output of this command i.e. 1544)
So 10,000,000 divided by 1,544 = 6,476 (the calculations don’t play with decimal values after the whole number)

Then take the delay value and divide by 10. So 40,000 becomes 4,000. This is because EIGRP calculates the delay in 10’s of microseconds rather than the microsecond value displayed on the output of this or the sh interface command.

Finally you have the figures your interested in, they are Bandwidth = 6,476 and you add that to the delay 4,000

Take that figure, shake it all about 😉 .. no don’t do that, simply take that figure of bandwidth (6,476) plus delay (4,000) and multiply it by 256 and your metric calculation will be displayed before your very eyes..
(6,476 + 4,000) x 256 = 2681856 which is your Feasible Distance to the network in question. 🙂

I hope that takes some of the pain out of understanding the metric for you peeps, I could have done with someone writing this as simply as this before I needed to understand it. 🙂
Be Cool.