Data ReHab

Allow me to apologise for not posting for such a long time. Embarrassingly, I'd become a packet sniffing junkie, however, I'm sober now. As part of my therapy I would like to come clean on my story to you.

It all started after a hard days work troubleshooting a particular network fault. Our edge router kept crashing randomly and we just couldn't pin it down. Morris suggested that we stay back after-hours to kick back with some beer, sniff a few packets and get high. It was a stressful day and being open to experimentation I agreed to try it out.

I didn't learn until later that Morris had been a casual packet sniffer from way back. He started sniffing X.25 frames back with his mates at uni. However, unlike me, he never let things get out of control.

I'll never forget the rush I felt as the hex of my very first sniffed packet injected itself onto the screen. However, the euphoria didn't last long. It never does and soon your buffer is empty and you need another hit. Just a small amount to keep you going. Keep you alert and on your toes.

You convince yourself to increase your buffer size and take a bugger sniff. Management didn't seem to mind. On the contrary, they were happy to see me working back late and taking such committed action. If only they knew the toll it was taking on my sole. Pretty soon the hit I was getting from sniffing traffic at the edge was wearing pretty shallow and I thought that the only way I could get satisfaction was if I started sniffing the big stuff. You know, some of the Gigabit Ethernet trunks at the core. At the peak of my illness, I would sniff about 2-5 Gigabytes worth of traffic at a time and stay at work all night analysing it. It was my artificial paradise.

After one mammoth session consisting of sniffing 40 Gig of Ethernet frames, 5 Meg of SNMP traps, some router debug, syslog and half a dozen or so router configs, my body had enough and I passed out on the keyboard. When I woke up I was in data rehab.

The hardest part was admitting that I had a problem. At the time I just thought I was a diligent engineer, who was perhaps just a little was more committed than most. However, after listening to some of the other guys in rehab I came to realise that I was just like them. This particular guy, "Ernie", had been in and out of data rehab a number of times. He was recently readmitted for turning on all the SNMP traps on all of his network gear. This quickly overdosed his network management software and almost killed himself and his network.

The most important thing that data rehab taught me is that you can't solve network problems by sniffing packets. Sniffing is fun, it will give you a buzz but is a destructive waste of your life. No matter how much or how little you sniff the problems going to still be there and you're going to want more

It's not that I don't get the craving for sniffing anymore. I do and it's still as strong as it ever was. Heck, Morris still dabbles in SNMP from time to time and I hear that SNMP V3 is big with the security folk. Its just that I know that I'm better off without it.

I now have the confidence to solve my network problems without having to ever sniff again. Debug clean for 45 days.

Claude

Join us for a game of D&D

I just wanted to note that myself (Claude), Morris and Horrace will be playing D&D on the Internet tonight. You are all welcome to join us. To clear up any confusion, this D&D stands for "Dungeons and Dragoons". It's a MMORPG set in the American War of Independence. My character (who has just acquired the 'wield dual muskets skill') is Sergeant at Arms, Timmothy Matheson of the Plymouth Militia. Our characters are currently immobile and face a tough challenge after our general failed his 'bypass pointless bureaucracy' roll.

There has been some request for a wide screen version of the CSMACD wall paper. We checked the source footage that Morris acquired and there is indeed more content sitting behind the black bars:

1280 x 854
1680 x 1050

Guatemalan Adventures & First Ever Photo of IP Packets

I have some very exciting news for you today that is simply mind-blowing. However, first I must apologise for the lack of updates in the very exciting world of the CSMACD.

It all started shortly before Christmas. I was on my way to Guatemala to install a voice over DECnet solution as part of a working holiday. Unfortunately the project fell through and I ended up having to pick coffee beans to afford my flight home.

Here is a photo of me standing on the edge of the valley where we have the coffee plantation:

No doubt you are waiting for the exciting news. Morris has managed to take the first ever live photos of IP packets. These photos were taken using an electron scanning microscope. The subject was a KupiTech layer 2.5 MLPS switch using CAT 7.5 UTT cable. The data itself is IP encapsulated CCS7MACD frames.

Some members of our team are in awe of the sheer majesty of the packets. Its for this reason that we have decided to make the ground-breaking photos available for download and commissioning them to your desktop.

1024 x 768
1280 x 960
1600 x 1200

Note: Desktop images are probably not compatible with the C64 GEOS environment...

Tear in the ATM Switching Fabric

A few weeks ago we experienced a large amount of dropped traffic over our main WAN link to the head office. With the help of Morris and our vendor, we isolated this to a faulty ATM fabric and swapped it out. However, it wasn't until today that we found out the real cause of the problem.

We sent the faulty fabric back to the vendor for further diagnosis. They believe that the fabric was worn down by faulty ATM cells.Eventually, the friction generated by the cells had ripped a hole right through the fabric itself. Data was quite literally being dropped out of the switch. Our vendor has sent us some photos of the torn fabric which they took using a electron microscope:

Figure 1: Normal Healthy ATM Fabric	Figure 2: Torn Faulty ATM Fabric

With the help of our vendors electron microscope we plan to study some of our ATM cells up close. Hopefully, the photos of the cells will show up the cause of the underlying problems. Until then, we just have to cross our fingers and hope that the new fabric holds together.

CAT 7.5 UTT Preview

Those of you who have been following the CSMACD closely will realise that CAT7.5 UTT forms an integral part of our IPoV solution. However, apart from the revolutionary 'parity' wire, most of the details of CAT7.5 have been kept a closely guarded secret.

That is until now...

Ask yourself this: "What has been sadly lacking in data cables to date?" Sure, your average cable is the right length and can be expected function well under most conditions. However, this they just don't deliver on one particular issue...

Networking cables taste horible!

The CSMACD Catagory 7.5 UTT cable comes in seven differnt flavours, enough to satify even the most fussy network engineer. We know that you are curious, so here is a little "taste" of things to come...

CAT7.5 UTT - Taste sample

(You will need a taste compatable browser - click here to search for a plugin)

IPoV Live Test - World First

Having implemented the below CCS7MACD network, today we established the worlds first IPoV call. The results were not a good as expected but we have decided to upload them to the public for general comment:

Download IPoV Test (MP3 - 1,152Kb)

What we require is some feedback from our pool of readers on how we can improve IPoV as a concept and even the CCS7MACD protocol. Please feel free to contribute and post your comments...

UTT In-Depth

The current development in the unshielded twisted triplet (UTT) cable is going well. There has been some confusion on how it is all structured, so I decided to put together this overview.

Standard unshielded twisted pair (UTP) has two conducting wires that send data:

Figure 1: Unshielded Twisted Pair

The pairs are twisted together; one carries a positive DC current, while the other carries a negative DC current. This helps neutralise the magnetic field created by the current, so that it does not induce current in other nearby pairs.

Unfortunately, the twisting and positive and negative currents do nothing to stop frame collisions. This is where UTT is superior:

Figure 2: Unshielded Twisted Triplet

By implementing a parity wire, this cable allows a signal to literally swerve off its current wire and avoid a fatal collision. Notice that the corners of the parity wire are banked and have extra width. This allows the signal to achieve drift.

UTT is still a prototype but investors have displayed keen interest and a contract for funding is under negotiation. We are looking at launching our first UTT category 7.5 cable at the end of next quarter.