50% growth for NetEvidence
To read the press release, download the pdf NetEvidence’s SaaS subscriptions increase by over 50 per cent

As a famous commentator once said “football is a game of two halves”. In one half, we have the event, the stadiums, the players and in the other we have the audience. However over 90% of the global audience will not be in the Rainbow nation for the next month – they’ll be glued to screens both big & small to get their fix of the 64 matches and highlights. This is the most connected World Cup ever – with the most methods of accessing the matches, from analogue & DAB Radio’s, to SmartPhones to Broadband to PCs at work. I think that fighting this 4yr event in 2010 in pointless – embrace it, and put in place facilities to allow people to see the matches therefore minimise the effect on Internet capacity and application slowdown. Businesses can use Highlight to view the affect when each match starts & stops, policing can help but the distraction won't go away. SKY’s Comment, Guardian Comment, Times Comment
My money is on _________ to win…. See you again in July when the dust settles.
Will Cisco’s delivery issues let other Vendors shine??
As the thaw begins in Enterprise & Providers IT budgets, project managers & backers are going to have to factor in an unexpected aspect – Cisco has been having issues with supply. I would imagine that this presents a window of opportunity for other vendors to steal a jump on the ‘big blue’. The last 18 months of spend freeze is now easing and project managers & IT organisations want the product now - Cisco has always been the low risk option but others continually shout about offering more functionality for the same or even less money… we’ll see how this situation plays out.
On the 26th March 2010, NetEvidence were delighted to be invited by Equinix to the launch party of their new Datacentre in Geneva – the 50th such facility globally. The event took place over 2 days at ‘The Globe” Exhibition Centre next to CERN on the Swiss / French border – home of the LHC (Large Hadron Collider). The event covered Green IT on day 1 & Cloud Solutions on day 2.


During the recent chaos caused by the volcanic ash cloud – what happened with use of collaborative technologies such as video conference and telepresence and of MSN Video & Skype video – business users at one end and domestic users at the other?
We estimate that usage of these technologies will have rocketed as users tried to keep talking face to face, but people are creatures of habit and once they get home & back to work will they continue to make use of the capability? For domestic people may have found a way of ‘talking’ that is preferable to the telephone. We think not – are we alone?
The dust is settling just a little on CA buying Nimsoft.
Hmm. On balance CA did the right thing, but they have an interesting integration challenge ahead. Of course, the stated intention is to leave Nimsoft alone – but the temptation to spread just a little of that pixie dust over their other products must be huge. eHealth – CA’s high-end reporting tool - is powerful and scalable, but we regularly meet customers put off by its complexity, and it’s growing long in the tooth. This is fine : it’s circle of life stuff, and plenty of large organisations have made major investments in the platform and will continue to buy it. But Nimsoft’s stated mantra of ‘unified monitoring’ must look awfully attractive when you’re trying to market an Application Reporting tool (Wily), an agent-based Flow tool (NetQos), an older Network Reporting tool (eHealth) and a handful of other acquisitions, all under the same brand.
The skill for CA will be in taking that pixie dust (basically, the energy and agility of a young company) and breathing deeply – without breaking what they just bought. But - we’ve seen Nimsoft as healthly competition for a few years, and wish them well ; CA’s acquisition is proof if it were needed that this segment is hot at the moment.
I’m still slightly dazed from attending and speaking at the recent Ethernet Summit conference in Paris.
Despite a technical bias, the lack of knowledge on the part of Mobile operators as they plough into the new world of IP backhauls is at first puzzling and then alarming. This really is worlds-collide stuff, judging by some of the questions being asked: mobile operators speaking a completely different language from the Ethernet providers trying to sell them backhaul. Ethernet represents a cost-saving opportunity for mobile but also a significant ceding of control to a third party and you can sense the reluctance.
There’s a very steep learning curve ahead as both sides strive to avoid repeating the iPhone mess (caused by 35 million of the gadgets trying to push data over cell networks that were designed for voice).
Last week we came across the perfect network.
I was sitting in on a Sales call with a Service Provider, and the subject of performance monitoring came up. We like to think we have this requirement covered pretty well, and our sales people were underlining the fact that Enterprises these days like to keep a close eye on network delay - in fact one of the major decision factors when an enterprise chooses a Network Provider is not just how fast the network is, but how easy it is for them to keep an eye on that performance.
"There's no point monitoring that with our network" explained a senior IT guy who worked for the Network Operator. "Network delay is fixed at 42 milliseconds ; it's part of the overall design".
Which is a little like saying there's no point locking your house when you go out, because there's a law against people stealing things. The whole point of monitoring networks is that (a) they do unexpected things and (b) even if they're behaving, you want a simple way to verify that. It's a good example of the 'counterflow' currently visible in the enterprise network market. While technology has traditionally flowed 'outward' - so that Service Providers find, recommend and sell new technology as it comes onto the market, for example - we're seeing that Service Requirements are starting to flow 'inwards'. Enterprise Customers are actually well ahead of their Network Operators in knowing what levels of service they want, and what level of visibility they'll demand. Clearly, a lot of Network Operators are having a hard time accepting this. The ones that anticipate and respond to this demand, will be the ones that win.
A shining example of why Service Levels get mis-sold, and no-one cares.
Meeting with a Service Provider last week, and their Account Team was explaining about a problem customer. "During the bid process, they demanded we provide continuous MOS monitoring on the network after it was installed," they explained, "to prove the quality and delay is within limits for voice traffic". MOS (Mean Opinion Score) is an old but simple scoring system for rating the quality of telephone calls. "Of course we agreed, but actually we've no way to provide the monitoring. So although we reckon the network is within limits, we've no way of proving it, so we're having to pay the customer £[a large amount] every month in compensation".
Not good news, you'd think, but the wonder of internal budgets provides the silver lining. Service Credits - that is, compensation for underperforming networks, including this one - are paid by another division within the Service Provider. It comes from another budget. And so the Account Team see no impact on their revenue and have no incentive to do anything about the problem. This is not good news for the Service Provider. Their customer will see them as lightweight and disorganised. Would you trust them with business-critical applications and traffic ? I'd pocket the refunds with a smile, and quietly recognise that long-term, they might not be the carrier for me. Food for thought.
You’re at a party and someone asks what you do. “Oh, I work in IT ” you say without thinking, “My company works with computer networks and the Internet”.
Big mistake. I realised recently why a friend of mine, a doctor, tells me that she never reveals her true profession at parties, since letting on you’re medical is a sure way to end up being quizzed on everyone’s latest aches, pains and bugs. And worryingly, being in Telecomms is going the same way. You see, to most of the general public being ‘In the Internet’ means you know everything, about every network. You know without further investigation why John’s broadband connection mysteriously dies every twenty minutes. You know exactly why Nancy can’t download attachments via her BT online account. You wrote Facebook and Twitter, without thinking of a decent business model, and you are personally responsible for most of the Spam on the internet. Oh, and most of all, you know Why the Internet Is So Slow.
This is heady stuff, and everyone wants to know the answer. Conversations about Starbucks, Holidays in Morocco and Finding a good Babysitter screech to a halt. “Oh baby – I have to hear the answer to this one”. You pause. The room goes quiet. From being a simple party guest you have suddenly become The Person who Runs the Internet. An air of menace begins to creep in.
Technical explanations rise to your lips – and die. These are not technical people. These are users, in their purest state and you’ve unwittingly wandered into a cage full of them. Talk of DLSAMs and Broadband Oversubscription, Ad-funded Business Models or Spambot technologies won’t work here. These people have been whipped to a frenzy by the Sunday Papers and Broadband Marketing, promising blindingly fast connections and a ‘rich multimedia experience’. Some of them really thought that an 8M broadband connection would be twice as fast as a 4M one. Some of them clicked on the attachment that came with that interesting email and found that Social Networking goes way, way beyond the stuff you see on MySpace. Reality hurts, especially when you realise using a Comparison Website to choose your broadband provider now means the Tech Support line has eight recorded messages between you and a human being, and they only want to talk about EastEnders in a bad accent or sell you an upgrade.
My advice : change professions at parties, at least temporarily. Feign ignorance, or feign bewilderment, or feign thirst and head for the kitchen. It usually works for me.
Some discussions with an industry magazine recently provoked the question of whether, and to what extent, automated tools can make life easier for network and IT managers. We've all seen the claims : automatic network discovery ; root cause analysis ; even artificial intelligence modules.
The really interesting thing is watching what happens in the Operations room (NOC) for a good-sized network, when something goes wrong. The engineers have never (at least while I've been watching) leapt onto a complex management platform like this and run root-cause analysis programs. What they do is pop up a console window onto the nearest large router, and start typing commands to check connectivity, route tables, and so on. This is how they think - they want a close, tight interface with what's going on within the network. Not a thick layer of automation in between.
Overall, a major risk with network managers is that they buy into this 'automated is good' vision and expect that they can install a piece of software which will continually crawl around their network, maintain a picture of what's there, and tell them what's broken. In practice, this doesn't work. Every data network (and 'data' is the keyword there - voice networks are very different) regardless of size is held together by a small core group of bright engineers who work very, very hard and carry much of the network 'state' in their heads. We see this in organisations of every size, from tiny companies to Fortune-100s and global service providers. It happens because data networks change too fast for any manual documentation to stabilise and be useful for more than about a week, and they exist in too complex and subtle a set of configurations for any automated system to be accurate and useful.
Which means organisations that want to improve discovery, documentation and problem recovery should not look to implement monolithic tools that claim to magically solve the problem, but rather should identify these core engineers and give them (perhaps simpler) tools which help them do their job better.