Trading-in: never accept the first offer
When upgrading network equipment, most Network Managers seek a trade-in on their old devices. There's absolutely no point in throwing away kit for which you could get a financial return, reducing the net cost of the upgrade.
That's fairly obvious, but there's more to this than meets the eye. The key is to shop around. Never accept the first trade-in price you are offered for any piece of redundant network equipment. It's almost certain to be well below what that piece of kit could command.
$50,000 saved
Several years ago, Go Communications had a customer who came to us periodically for devices that their incumbent supplier, a manufacturer, could not deliver quickly enough. They had reached the point where they needed to upgrade their network, in the process removing some 120 48-port VoIP-enabled switches. Their manufacturer offered them a trade-in of $200 per switch - around one twentieth of their purchase price some two-to-three years earlier.
Understandably, the customer felt they should be able to do better, and approached us. We were able to offer them around $600 per switch - a total saving on the upgrade of almost $50,000 (around £30,000). We were able to do this as VoIP had recently become a key technology for many up and coming businesses, and there was a strong second-user market for these VoIP enabled switches. The manufacturer could not match our offer because selling second-user equipment is simply not their core business.
Three rules
This story illustrates three simple rules of network equipment trade-ins. Firstly, always shop around. Secondly, never think you can second-guess the market. Thirdly, manufacturers are always likely to offer you a much lower price for your old hardware than dealers with orders to fulfil for those products. It's astonishing how many businesses throw away huge sums byfailing to follow these three rules.
Let's take a closer look at that third point. Switches, routers and other network devices are almost always deployed with a variety of options installed - extra features, modules and software functionality, for example. When you ask your manufacturer for a trade-in value for your redundant devices, the chances are, planning on feeding those devices back into their refurbishment programme, that they will offer you a price based on their base configurations. No options, no additional modules and no extra software functionality.
Additionally, handling your old kit, refurbishing it and selling it on is a headache for the manufacturer. Their core business and chief objective is to sell their current lines, not to resell old equipment. Really, they'd rather you simply disposed of your redundant equipment than traded it in. On top of all that, any offer they make you for it will come straight off their bottom line: it's likely to be grudging at best.
A seller's market
Now, consider the dealer who needs your old equipment to fulfil a customer order. The scenario is turned on its head: without your kit that dealer may not have a bottom line on that order at all. They have an active, pressing need for your equipment - you're in a seller's market, and can command a much higher price.
There's a direct parallel to all this that most of us are familiar with. When buying a new car, accepting the dealer's trade-in offer on the old one is convenient, easy and simple, but we all know that we'll get much more for it by selling it privately. Only by doing so can we make sure the value of all the options - that heated, massaging driver's seat, the iPod port, those parking sensors - is reflected in the price we get for it.
So - especially in the current economic climate, with cost control a priority and used equipment prices on the rise as a result of fewer organisations upgrading - swim against the tide. Never, ever accept the first trade-in price you're offered on your old networking kit. Always shop around. It could save you literally tens of thousands of pounds
Tuesday 17 November 2009
Thursday 1 October 2009
Seek a second opinion
Don't be a lone ranger
One of the fascinating things about networks is that, to achieve any given desired end result, there is always a virtually limitless number of possible approaches. That makes seeking a second opinion on your proposed solution to any given challenge an absolute necessity.
Whether you need to upgrade, replace, modify, extend or reconfigure your network, there will always be many different possible ways of doing it. Only one of them will be the best approach for your particular needs. Tackling the problem of nailing that one down on your own is a tough challenge.
I love challenges like this - they are like a complex puzzle that needs to be solved. And they come up frequently, in all sorts of circumstances. Maybe a solution designed for a particular scenario doesn’t work - perhaps the details from the original network starting point aren’t right, there's been a change of WAN connection or a last-minute change of software that the hardware doesn’t support. Or maybe there's less budget or physical space than had been expected.
In such situations, changes must be made, and selecting the best changes is made vastly quicker, easier and more effective by securing an expert second opinion. Viewing the situation in a different light and from a different perspective usually turns up unexpected options that would otherwise be overlooked. And those options can sometimes save thousands of pounds and untold effort.
Working from those options, you can then create lists of the components needed for each one, and cost them up. One of those quotes will be lower than the others, for the same end result, and there's a good chance it will be one that you hadn't previously considered.
For this reason, whenever I propose a solution I always offer at least two alternatives for the customer to consider - this is particularly important in the current economic climate. For example I will often offer two single-port modules in place of a single, more popular dual-port module. The laws of supply and demand dictate that the former is nearly always cheaper than the latter.
This approach is particularly invaluable in situations where a proposed network solution is beyond the scope of your budget. An expert second opinion can almost always deliver canny ways of removing cost from a quote, potentially using products from completely different ranges. This is especially so when a network design has been over-specified in an attempt to cover all eventualities - I can almost guarantee that there will be a cheaper alternative just waiting to be found by a fresh pair of eyes.
Don't go it alone. Always seek a second opinion.
One of the fascinating things about networks is that, to achieve any given desired end result, there is always a virtually limitless number of possible approaches. That makes seeking a second opinion on your proposed solution to any given challenge an absolute necessity.
Whether you need to upgrade, replace, modify, extend or reconfigure your network, there will always be many different possible ways of doing it. Only one of them will be the best approach for your particular needs. Tackling the problem of nailing that one down on your own is a tough challenge.
I love challenges like this - they are like a complex puzzle that needs to be solved. And they come up frequently, in all sorts of circumstances. Maybe a solution designed for a particular scenario doesn’t work - perhaps the details from the original network starting point aren’t right, there's been a change of WAN connection or a last-minute change of software that the hardware doesn’t support. Or maybe there's less budget or physical space than had been expected.
In such situations, changes must be made, and selecting the best changes is made vastly quicker, easier and more effective by securing an expert second opinion. Viewing the situation in a different light and from a different perspective usually turns up unexpected options that would otherwise be overlooked. And those options can sometimes save thousands of pounds and untold effort.
Working from those options, you can then create lists of the components needed for each one, and cost them up. One of those quotes will be lower than the others, for the same end result, and there's a good chance it will be one that you hadn't previously considered.
For this reason, whenever I propose a solution I always offer at least two alternatives for the customer to consider - this is particularly important in the current economic climate. For example I will often offer two single-port modules in place of a single, more popular dual-port module. The laws of supply and demand dictate that the former is nearly always cheaper than the latter.
This approach is particularly invaluable in situations where a proposed network solution is beyond the scope of your budget. An expert second opinion can almost always deliver canny ways of removing cost from a quote, potentially using products from completely different ranges. This is especially so when a network design has been over-specified in an attempt to cover all eventualities - I can almost guarantee that there will be a cheaper alternative just waiting to be found by a fresh pair of eyes.
Don't go it alone. Always seek a second opinion.
Wednesday 2 September 2009
Change your network manufacturer
Time for a change?
All network systems manufacturers are not the same. While this may not be a startling revelation, a key consequence does often go overlooked: changing your network manufacturer, or even simply considering doing so, can save you significant amounts of money.
Different manufacturers invest different amounts of effort in technology research and development. As a result, your current network systems vendor may well have missed a trick or two on the platforms you use.
Upgrade or replace?
If your current network systems have been in place for some time and you are considering upgrading to a newer platform from your existing manufacturer, it's worth casting your net a little wider. It may be less expensive to replace your entire network, or a segment of it, with one from an alternative manufacturer.
The savings can go further still. By switching manufacturers you can drive down not only your equipment purchase cost, but also your ongoing costs. As well as providing additional and enhanced features, newer technologies can also deliver reductions in hardware size, power consumption, and cooling and maintenance requirements. All of these translate into reduced costs, and reduced carbon footprint, which can be a key PR benefit for your business.
It's clear that all of these savings can be realised by switching to a more appropriate network systems manufacturer. What is perhaps not so obvious is that significant savings can be secured without moving at all.
Chaos on the network
Picture the scene. Part of your network is down. Maybe a power supply has failed, or perhaps a card has died. Either way one of your main routers is out of action. Traffic is being diverted over the rest of the network but, nonetheless, the entire network is running well below its normal speed. All manner of unexpected knock-on problems are arising. Web access is painfully slow, for example, and emails are not getting through. Inevitably, the phones are ringing constantly, with disgruntled and stressed users on the other end.
Into the eye of this storm drops a call from your network systems manufacturer. It's your account manager, chasing his monthly numbers: "What are you ordering this month? Where are you buying it from? Can I have a list of what you bought last month please? Can you email me the details today please?" You find yourself wondering who the customer is in this scenario, and whether there may be a bit of tail-wagging-the-dog going on.
Proactive, not pestering
Now, distant, detached network systems suppliers are of little use. Manufacturers design and build their products for specific tasks, so accurate, detailed and up to date product knowledge is key to successful network device specification. I am all for a proactive approach to assisting and supporting customers – helping them build the network most suitable for their needs.
There is a thin dividing line, though, between proactivity and pressure selling, with constant focus on sales numbers and pestering for orders. The key to making sure your suppliers stay on the right side of this line is to ensure that they always know that you have the option of taking your business elsewhere.
My very first network systems customer, back in the 90s, was an expert at this. He kept all his suppliers on their toes, and at the first sign of any of them trying to push him, would make it quite clear that if they didn't back off, he would be recommending a switch of suppliers at the next network refresh. It never failed. Manufacturers should work hard to win your business, but in the right way. Don't let complacency set in. Make sure your suppliers stay absolutely clear on who the customer is. They will quickly get the message that you won't be at their beck and call, and that you are likely to be the sort of customer who will very carefully scrutinise quotations when it comes to ordering new equipment and services.
All network systems manufacturers are not the same. While this may not be a startling revelation, a key consequence does often go overlooked: changing your network manufacturer, or even simply considering doing so, can save you significant amounts of money.
Different manufacturers invest different amounts of effort in technology research and development. As a result, your current network systems vendor may well have missed a trick or two on the platforms you use.
Upgrade or replace?
If your current network systems have been in place for some time and you are considering upgrading to a newer platform from your existing manufacturer, it's worth casting your net a little wider. It may be less expensive to replace your entire network, or a segment of it, with one from an alternative manufacturer.
The savings can go further still. By switching manufacturers you can drive down not only your equipment purchase cost, but also your ongoing costs. As well as providing additional and enhanced features, newer technologies can also deliver reductions in hardware size, power consumption, and cooling and maintenance requirements. All of these translate into reduced costs, and reduced carbon footprint, which can be a key PR benefit for your business.
It's clear that all of these savings can be realised by switching to a more appropriate network systems manufacturer. What is perhaps not so obvious is that significant savings can be secured without moving at all.
Chaos on the network
Picture the scene. Part of your network is down. Maybe a power supply has failed, or perhaps a card has died. Either way one of your main routers is out of action. Traffic is being diverted over the rest of the network but, nonetheless, the entire network is running well below its normal speed. All manner of unexpected knock-on problems are arising. Web access is painfully slow, for example, and emails are not getting through. Inevitably, the phones are ringing constantly, with disgruntled and stressed users on the other end.
Into the eye of this storm drops a call from your network systems manufacturer. It's your account manager, chasing his monthly numbers: "What are you ordering this month? Where are you buying it from? Can I have a list of what you bought last month please? Can you email me the details today please?" You find yourself wondering who the customer is in this scenario, and whether there may be a bit of tail-wagging-the-dog going on.
Proactive, not pestering
Now, distant, detached network systems suppliers are of little use. Manufacturers design and build their products for specific tasks, so accurate, detailed and up to date product knowledge is key to successful network device specification. I am all for a proactive approach to assisting and supporting customers – helping them build the network most suitable for their needs.
There is a thin dividing line, though, between proactivity and pressure selling, with constant focus on sales numbers and pestering for orders. The key to making sure your suppliers stay on the right side of this line is to ensure that they always know that you have the option of taking your business elsewhere.
My very first network systems customer, back in the 90s, was an expert at this. He kept all his suppliers on their toes, and at the first sign of any of them trying to push him, would make it quite clear that if they didn't back off, he would be recommending a switch of suppliers at the next network refresh. It never failed. Manufacturers should work hard to win your business, but in the right way. Don't let complacency set in. Make sure your suppliers stay absolutely clear on who the customer is. They will quickly get the message that you won't be at their beck and call, and that you are likely to be the sort of customer who will very carefully scrutinise quotations when it comes to ordering new equipment and services.
Monday 20 July 2009
Have kit ready for plug and play
Ship ready for 'plug and play'
We've all heard it, or something very like it, a hundred times. "What time will you be home? The kids have forgotten what you look like." Anyone working with networks is familiar with long and unpredictable hours - in part because, let's be honest, we're all pretty much hooked on what we do.
What we do is fun. At least, it's fun when you can log in and out, engage and disengage, when you want to. It's less fun when you have to revisit work that's been done before and shouldn't need to be redone. For my part, I'd much rather be home reading my son "Slinky Malinky" than rehashing work that really should have been put to bed already. Life's too short.
Life is too short, and on top of that, the economy is too unforgiving. Time, as has been noted so often, is money. Your engineers' time is no different to anyone else's in this respect, and wasting it is a direct cost to your business. In the current economic climate, with growth and even survival depending so heavily on cash flow and profit, that cost could make the difference between staying afloat and going to the wall. Making cost savings without impacting quality is a key priority.
Build, configure and test
Let's look at a common scenario. A customer's network installation is just around the corner - you're waiting for delivery of the network devices and associated items. Engineering resources have been planned and allocated. The customer is champing at the bit and the pressure is on.
What happens next? You receive the equipment, unpack, build, configure and soak test it. A couple of days later, you dismantle it, pack it up again and ship it off to the customer.
A simple question
At this point, I'd like to ask a very simple question. Why do it this way? On receiving the equipment the customer or your installation engineer will unpack, build, install, configure and test it. There is more than a little duplication of effort going on here. There's increased risk of damage, and time and money are being wasted.
Avoiding this risk and waste is actually very easy. While awaiting delivery of the network devices, memory, cables and software, call the customer and ask them if they would like the equipment supplied with their configuration pre-loaded, with any necessary memory and software upgrades made. In short, would they like it delivered ready for plug and play installation? I can almost guarantee that most will give you an emphatic "Yes please".
Make life easier
This will make the customer's life easier, of course, but it will also make your life easier and save you time and money. Keeping the equipment in your testing station until the memory and software has arrived, and then making the necessary upgrades there and then will mean the job is done and the equipment can be shipped to the customer ready to run, with a test report certifying the upgrades that have been made.
As well as saving you time and money, this approach avoids the situation in which the customer discovers that a device has insufficient memory for the required software image. It also means that the customer gets network equipment that has been tested as they will use it - and that's the test that really matters.
Investing this small amount of additional effort before shipping the equipment to the customer can save your installation engineers hours of on-site time. That means savings in engineering costs, improved customer satisfaction, fewer network teething troubles, and, of course, that you and your engineers get to remind yourselves of what your family look like. That's what I call a result.
We've all heard it, or something very like it, a hundred times. "What time will you be home? The kids have forgotten what you look like." Anyone working with networks is familiar with long and unpredictable hours - in part because, let's be honest, we're all pretty much hooked on what we do.
What we do is fun. At least, it's fun when you can log in and out, engage and disengage, when you want to. It's less fun when you have to revisit work that's been done before and shouldn't need to be redone. For my part, I'd much rather be home reading my son "Slinky Malinky" than rehashing work that really should have been put to bed already. Life's too short.
Life is too short, and on top of that, the economy is too unforgiving. Time, as has been noted so often, is money. Your engineers' time is no different to anyone else's in this respect, and wasting it is a direct cost to your business. In the current economic climate, with growth and even survival depending so heavily on cash flow and profit, that cost could make the difference between staying afloat and going to the wall. Making cost savings without impacting quality is a key priority.
Build, configure and test
Let's look at a common scenario. A customer's network installation is just around the corner - you're waiting for delivery of the network devices and associated items. Engineering resources have been planned and allocated. The customer is champing at the bit and the pressure is on.
What happens next? You receive the equipment, unpack, build, configure and soak test it. A couple of days later, you dismantle it, pack it up again and ship it off to the customer.
A simple question
At this point, I'd like to ask a very simple question. Why do it this way? On receiving the equipment the customer or your installation engineer will unpack, build, install, configure and test it. There is more than a little duplication of effort going on here. There's increased risk of damage, and time and money are being wasted.
Avoiding this risk and waste is actually very easy. While awaiting delivery of the network devices, memory, cables and software, call the customer and ask them if they would like the equipment supplied with their configuration pre-loaded, with any necessary memory and software upgrades made. In short, would they like it delivered ready for plug and play installation? I can almost guarantee that most will give you an emphatic "Yes please".
Make life easier
This will make the customer's life easier, of course, but it will also make your life easier and save you time and money. Keeping the equipment in your testing station until the memory and software has arrived, and then making the necessary upgrades there and then will mean the job is done and the equipment can be shipped to the customer ready to run, with a test report certifying the upgrades that have been made.
As well as saving you time and money, this approach avoids the situation in which the customer discovers that a device has insufficient memory for the required software image. It also means that the customer gets network equipment that has been tested as they will use it - and that's the test that really matters.
Investing this small amount of additional effort before shipping the equipment to the customer can save your installation engineers hours of on-site time. That means savings in engineering costs, improved customer satisfaction, fewer network teething troubles, and, of course, that you and your engineers get to remind yourselves of what your family look like. That's what I call a result.
Wednesday 24 June 2009
Buy 2 spares for mission critical networks
Hardware maintenance: a better way
What do you do when a critical device fails on your network? You call your on site maintenance company, they send an engineer with a replacement unit, and you're back up and running, right? You do have a maintenance contract covering your mission-critical network devices, don't you?
Actually, you may not need such a contract. And if you have one, you could well be throwing thousands of pounds away every year, completely unnecessarily, in fees. That may well sound like complete lunacy, but stick with me: there's good sense behind it.
Firstly, let me say that if your network is mission critical, and you don't have network engineers on your team, then you almost certainly do need to buy in maintenance support – or get network engineering expertise on your payroll.
Spare up!
If you have your own engineers, though, you can make significant savings by keeping refurbished spares on hand to cover all your critical network devices. In the event of a failure, your own engineers simply use the spare unit to replace the failed device, and the network is up and running again.
The maintenance contract fees for any given switch or router can be double the cost of a refurbished device. That means that if you hold two refurbished spares for each device on your network, your support costs in the first year will be no greater than they would have been with a support contract, and then for every subsequent year, your only expenses will be the replacement of any spare units used to replace failed units.
Over the life of the network, such savings can add up to very significant amounts.
You may be concerned at the prospect of using refurbished units rather than new. Such concern is unfounded, though: this is exactly what maintenance companies do. They know that there is little to choose between new and refurbished units when it comes to reliability, and on the cost front, the issue is a genuine 'no brainer' – the cost of a refurbished switch or router can be as little as one tenth that of a new unit.
Radical but not unprecedented
While this 'no maintenance contracts' strategy may sound radical, it is by no means unprecedented. I gather from a close friend in the UK's largest telecoms company that one of the City of London's largest and most respected investment banks adopted the approach some time ago. They do not use any external network maintenance company, and only use their network devices' manufacturer for software updates and TAC access.
With sufficient engineering resource on their payroll, entrusting the support of their network to a third party makes no sense – doing so would drive up both costs and risk.
This investment bank has done the job properly: they keep not one or even two but three disaster recovery systems in place. By doing so, they ensure they are proofed against almost any imaginable hardware failure, as well as other potential disasters such as terrorist attack. Even with this level of redundancy in place, they still make significant savings, which they plough back into the latest technology for the core of their network.
Companies that do things the same way that they have always been done tend to be the also-rans. Companies that innovate, finding new and better ways of operating, tend to be the high-flyers and the market-leaders. Indeed, in the current environment of deepening economic gloom, with tightening budgets and contracting markets, survival itself may well depend on such innovation.
What do you do when a critical device fails on your network? You call your on site maintenance company, they send an engineer with a replacement unit, and you're back up and running, right? You do have a maintenance contract covering your mission-critical network devices, don't you?
Actually, you may not need such a contract. And if you have one, you could well be throwing thousands of pounds away every year, completely unnecessarily, in fees. That may well sound like complete lunacy, but stick with me: there's good sense behind it.
Firstly, let me say that if your network is mission critical, and you don't have network engineers on your team, then you almost certainly do need to buy in maintenance support – or get network engineering expertise on your payroll.
Spare up!
If you have your own engineers, though, you can make significant savings by keeping refurbished spares on hand to cover all your critical network devices. In the event of a failure, your own engineers simply use the spare unit to replace the failed device, and the network is up and running again.
The maintenance contract fees for any given switch or router can be double the cost of a refurbished device. That means that if you hold two refurbished spares for each device on your network, your support costs in the first year will be no greater than they would have been with a support contract, and then for every subsequent year, your only expenses will be the replacement of any spare units used to replace failed units.
Over the life of the network, such savings can add up to very significant amounts.
You may be concerned at the prospect of using refurbished units rather than new. Such concern is unfounded, though: this is exactly what maintenance companies do. They know that there is little to choose between new and refurbished units when it comes to reliability, and on the cost front, the issue is a genuine 'no brainer' – the cost of a refurbished switch or router can be as little as one tenth that of a new unit.
Radical but not unprecedented
While this 'no maintenance contracts' strategy may sound radical, it is by no means unprecedented. I gather from a close friend in the UK's largest telecoms company that one of the City of London's largest and most respected investment banks adopted the approach some time ago. They do not use any external network maintenance company, and only use their network devices' manufacturer for software updates and TAC access.
With sufficient engineering resource on their payroll, entrusting the support of their network to a third party makes no sense – doing so would drive up both costs and risk.
This investment bank has done the job properly: they keep not one or even two but three disaster recovery systems in place. By doing so, they ensure they are proofed against almost any imaginable hardware failure, as well as other potential disasters such as terrorist attack. Even with this level of redundancy in place, they still make significant savings, which they plough back into the latest technology for the core of their network.
Companies that do things the same way that they have always been done tend to be the also-rans. Companies that innovate, finding new and better ways of operating, tend to be the high-flyers and the market-leaders. Indeed, in the current environment of deepening economic gloom, with tightening budgets and contracting markets, survival itself may well depend on such innovation.
Monday 18 May 2009
Don’t buy Future Proof
So you think you can predict the future?
Future proof. Future proof. Future proof.
It's almost a mantra in the IT world. Everyone is preaching it. Everyone is doing it. And hardware manufacturers love it. It guarantees larger sales, right now. And as long as customers continue to do it, it will guarantee larger sales tomorrow, too.
The thing is, while it works like a dream as a sales improvement tool for the hardware manufacturers, it doesn't generally work for the customer, except over the short term. In fact, I'd go so far as to say it's usually throwing money away on scalability that will never be used.
A huge future proofing over-spend
Part of Go Communications' work is in the decommissioning of existing network infrastructure, usually to make way for new systems. A while back I was involved in the decommissioning of a truly enormous network in a secret underground bunker somewhere in the South of England.
It was all a bit James Bond. There were metal detectors and a personal search on the way in, and a truly enormous switch and server farm inside. I had never seen an installation of such a size, before or since.
The decommissioning was required because the hosting company had overspent and gone out of business as a result. The customer it was designed for had required that the specification be future proof not just for a couple of years but for ten.
The waste was horrifying. Rack upon rack of slots sat empty in every single chassis and the cabinets were twice the height they needed to be. The crippling cost of the over-specified hardware and the ongoing costs of wasted electricity – larger cabinets need larger power supplies and more cooling – must have been contributors to the hosting company's cash-flow problems. They could even have been the issues that brought it down.
How far ahead can you see?
Future proofing sounds good. And I'm the first to agree that, over the short term – a year or two – it is a vitally important issue in most network designs. Trying to cater for a future more than a couple of years away, though, is almost guaranteed to fail, because it's virtually certain that the network will have been replaced before the expansion space has been used.
Coming back to the secret bunker switch and server farm, specifying for ten years ahead seems, on the face of it, to be prudent – especially with a large, mission-critical network. Let's give it a moment's cold, hard thought, though. Ten years ago, we were all using Windows 98. The original iPod was still more than two years away. Digital cameras were an exciting new thing. And the average desktop PC was less powerful than today's iPhones.
With technological advancement accelerating, the changes over the next ten years are likely to be even more dramatic than those over the last. Predicting developments in technology is a bit like forecasting the weather: if you know what you're doing, you can make reasonable guesses as to what's around the corner, but try doing it any distance into the future and you're setting yourself up for a fall.
When you're specifying network systems, a fall means expense, and usually a lot of it.
Plan ahead, but control costs
Specifying one or two extra slots for extra cards in the next few months or year is usually prudent, but it's almost always wasteful going any further. While there is always the chance that additional slots over and above those one or two may be needed, designing for that small possibility is rarely wise: it ties up money that could be much more effectively deployed elsewhere on the network today.
Specify what you need today, with a little expansion room for the immediate future. If you do find at some point in the future that you need extra network resources over and above those for which you've allowed space, then re-specifying the appropriate parts of the network to cater for those resources can usually be done economically, especially as the general trend of hardware prices is always downwards rather than upwards.
Future proof. Future proof. Future proof.
It's almost a mantra in the IT world. Everyone is preaching it. Everyone is doing it. And hardware manufacturers love it. It guarantees larger sales, right now. And as long as customers continue to do it, it will guarantee larger sales tomorrow, too.
The thing is, while it works like a dream as a sales improvement tool for the hardware manufacturers, it doesn't generally work for the customer, except over the short term. In fact, I'd go so far as to say it's usually throwing money away on scalability that will never be used.
A huge future proofing over-spend
Part of Go Communications' work is in the decommissioning of existing network infrastructure, usually to make way for new systems. A while back I was involved in the decommissioning of a truly enormous network in a secret underground bunker somewhere in the South of England.
It was all a bit James Bond. There were metal detectors and a personal search on the way in, and a truly enormous switch and server farm inside. I had never seen an installation of such a size, before or since.
The decommissioning was required because the hosting company had overspent and gone out of business as a result. The customer it was designed for had required that the specification be future proof not just for a couple of years but for ten.
The waste was horrifying. Rack upon rack of slots sat empty in every single chassis and the cabinets were twice the height they needed to be. The crippling cost of the over-specified hardware and the ongoing costs of wasted electricity – larger cabinets need larger power supplies and more cooling – must have been contributors to the hosting company's cash-flow problems. They could even have been the issues that brought it down.
How far ahead can you see?
Future proofing sounds good. And I'm the first to agree that, over the short term – a year or two – it is a vitally important issue in most network designs. Trying to cater for a future more than a couple of years away, though, is almost guaranteed to fail, because it's virtually certain that the network will have been replaced before the expansion space has been used.
Coming back to the secret bunker switch and server farm, specifying for ten years ahead seems, on the face of it, to be prudent – especially with a large, mission-critical network. Let's give it a moment's cold, hard thought, though. Ten years ago, we were all using Windows 98. The original iPod was still more than two years away. Digital cameras were an exciting new thing. And the average desktop PC was less powerful than today's iPhones.
With technological advancement accelerating, the changes over the next ten years are likely to be even more dramatic than those over the last. Predicting developments in technology is a bit like forecasting the weather: if you know what you're doing, you can make reasonable guesses as to what's around the corner, but try doing it any distance into the future and you're setting yourself up for a fall.
When you're specifying network systems, a fall means expense, and usually a lot of it.
Plan ahead, but control costs
Specifying one or two extra slots for extra cards in the next few months or year is usually prudent, but it's almost always wasteful going any further. While there is always the chance that additional slots over and above those one or two may be needed, designing for that small possibility is rarely wise: it ties up money that could be much more effectively deployed elsewhere on the network today.
Specify what you need today, with a little expansion room for the immediate future. If you do find at some point in the future that you need extra network resources over and above those for which you've allowed space, then re-specifying the appropriate parts of the network to cater for those resources can usually be done economically, especially as the general trend of hardware prices is always downwards rather than upwards.
Monday 20 April 2009
Don't specify new products too early
Caution: new products ahead
There's a certain amount of kudos in being 'at the leading edge'. Certainly, it's fun, especially if you have an engineering head on your shoulders, to try out the latest products. But, and this may come as something of a surprise from someone who makes his living selling networking products, ordering products before they are established in the market is not always a good idea. In fact, it can be a very bad idea indeed, and can cost your business dearly.
The leading edge is, of course, also known as the bleeding edge – and with good reason. Manufacturers almost always encounter unexpected design, production and logistical challenges with new product lines. In fact, over the past ten years, I can't remember a single network product that in its first release didn’t have some problem or other.
Recently, one of our customers specified the latest network switch from a particular manufacturer's range. Following a week of perfect, uninterrupted operation, the device spontaneously rebooted itself for no apparent reason. Brand new to the market, the switch had no field history: in specifying it the customer had taken a step into the unknown, with predictable results. Thankfully, they had the time and resources to handle the issue, but not all businesses would.
Resolved, but when?
Many bugs and issues with new products are of course resolved, although, when they first arise, no one knows how long such resolutions will remain in the pipeline. Worse yet, some are never resolved. Work arounds are almost always offered, but, again, no one can say how long devising them will take. In the meantime, who knows how many hours, days and weeks will have been lost?
In this respect, network switches, routers and other devices are like cars, washing machines, PCs and a host of other devices. Although a product's name, appearance and specifications may not change much over the years, a constant stream of improvements is made during its lifetime. Unless you ask, you don't generally hear about such improvements, but because of them the experience – and cost – of using the product some months after its launch can differ enormously from that of using it at Day One.
So here's my advice: unless you absolutely must, don't specify new products as soon as they are announced, or even released. Give them time to bed down in the market and for the inevitable teething problems to present themselves and be resolved.
Significant hidden costs
Of course, there may be compelling reasons for you to specify a brand new product – product features essential to your application and not available elsewhere, for example. If not, though, don't risk becoming a manufacturer's unpaid product tester, especially if you are pressed for time or have limited budgets, a small network team or a mission-critical network. Let your competitors discover the as-yet undiscovered bugs and handle the yet to be issued field notices – and bear the cost.
That cost, although often hidden, can be significant. In all aspects of business, time is one of the most valuable commodities you have at your disposal. And wasting it is one of the most expensive mistakes you can make. Just as jumping on the wrong bandwagon can cost you dear in terms of wasted time, so can specifying an unproven new product.
There's a certain amount of kudos in being 'at the leading edge'. Certainly, it's fun, especially if you have an engineering head on your shoulders, to try out the latest products. But, and this may come as something of a surprise from someone who makes his living selling networking products, ordering products before they are established in the market is not always a good idea. In fact, it can be a very bad idea indeed, and can cost your business dearly.
The leading edge is, of course, also known as the bleeding edge – and with good reason. Manufacturers almost always encounter unexpected design, production and logistical challenges with new product lines. In fact, over the past ten years, I can't remember a single network product that in its first release didn’t have some problem or other.
Recently, one of our customers specified the latest network switch from a particular manufacturer's range. Following a week of perfect, uninterrupted operation, the device spontaneously rebooted itself for no apparent reason. Brand new to the market, the switch had no field history: in specifying it the customer had taken a step into the unknown, with predictable results. Thankfully, they had the time and resources to handle the issue, but not all businesses would.
Resolved, but when?
Many bugs and issues with new products are of course resolved, although, when they first arise, no one knows how long such resolutions will remain in the pipeline. Worse yet, some are never resolved. Work arounds are almost always offered, but, again, no one can say how long devising them will take. In the meantime, who knows how many hours, days and weeks will have been lost?
In this respect, network switches, routers and other devices are like cars, washing machines, PCs and a host of other devices. Although a product's name, appearance and specifications may not change much over the years, a constant stream of improvements is made during its lifetime. Unless you ask, you don't generally hear about such improvements, but because of them the experience – and cost – of using the product some months after its launch can differ enormously from that of using it at Day One.
So here's my advice: unless you absolutely must, don't specify new products as soon as they are announced, or even released. Give them time to bed down in the market and for the inevitable teething problems to present themselves and be resolved.
Significant hidden costs
Of course, there may be compelling reasons for you to specify a brand new product – product features essential to your application and not available elsewhere, for example. If not, though, don't risk becoming a manufacturer's unpaid product tester, especially if you are pressed for time or have limited budgets, a small network team or a mission-critical network. Let your competitors discover the as-yet undiscovered bugs and handle the yet to be issued field notices – and bear the cost.
That cost, although often hidden, can be significant. In all aspects of business, time is one of the most valuable commodities you have at your disposal. And wasting it is one of the most expensive mistakes you can make. Just as jumping on the wrong bandwagon can cost you dear in terms of wasted time, so can specifying an unproven new product.
Friday 27 March 2009
Use manufacturer approved memory products
Genuine manufacturers' memory products. Wonderful, no doubt, but, let's be honest, they do tend to be a bit pricey.
You wouldn't buy a new TV from the high street if you could buy the same model for less than half the price from a reputable online store, would you? So why pay over the odds on your network switches and routers by using genuine manufacturers' memory products? After all, there are plenty of third party memory modules on sale for routers and switches.
Making savings
The obvious answer, of course, is that using third party memory would invalidate your warranty, wouldn't it? Or even not work at all. Well, no, in fact. It's something of a trade secret, but most manufacturers publish lists of approved third party memory products. And in the current economic climate, it's good commercial sense to find out where you could be making savings.
Those savings are far from insignificant, too. Approved third party memory product prices are on average some 80% lower than original manufacturers'.
The 'good old days'
Saving money on memory is nothing new of course. A few years back it was common for savvy network engineers to populate their live routers with third party memory, but keep one set on the shelf with original memory installed. That way they could have the benefit of the savings but avoid being caught out with a failed router populated with non-approved memory.
Times have changed, though. If, five or ten years ago, a router went down one evening, the engineer would typically not need to set about sorting it out it the next morning. Today, he's going to get a call-out when the router fails, and he's going to have to drop everything to see to it right away. His curry evening goes out of the window, his korma congeals and it's not going to help his heartburn. Really, it's not worth it.
No more heartburn
So here's what to do. Research your router and switch manufacturers' memory path upgrade programmes for each and every product you use. Find out who those manufacturers use as approved memory manufacturers. Most manufacturers publish the names, but they don't advertise them on their home pages: usually they are buried deep within their technical pages. If you cant find the information you're looking for, it's worth asking for it. You're quite within your rights to do so.
However you get to the information, once you have it, you've found the pot of gold at the end of the rainbow: safe memory upgrades for your sensitive network devices, no warranties invalidated, very significant savings. And no heartburn.
You wouldn't buy a new TV from the high street if you could buy the same model for less than half the price from a reputable online store, would you? So why pay over the odds on your network switches and routers by using genuine manufacturers' memory products? After all, there are plenty of third party memory modules on sale for routers and switches.
Making savings
The obvious answer, of course, is that using third party memory would invalidate your warranty, wouldn't it? Or even not work at all. Well, no, in fact. It's something of a trade secret, but most manufacturers publish lists of approved third party memory products. And in the current economic climate, it's good commercial sense to find out where you could be making savings.
Those savings are far from insignificant, too. Approved third party memory product prices are on average some 80% lower than original manufacturers'.
The 'good old days'
Saving money on memory is nothing new of course. A few years back it was common for savvy network engineers to populate their live routers with third party memory, but keep one set on the shelf with original memory installed. That way they could have the benefit of the savings but avoid being caught out with a failed router populated with non-approved memory.
Times have changed, though. If, five or ten years ago, a router went down one evening, the engineer would typically not need to set about sorting it out it the next morning. Today, he's going to get a call-out when the router fails, and he's going to have to drop everything to see to it right away. His curry evening goes out of the window, his korma congeals and it's not going to help his heartburn. Really, it's not worth it.
No more heartburn
So here's what to do. Research your router and switch manufacturers' memory path upgrade programmes for each and every product you use. Find out who those manufacturers use as approved memory manufacturers. Most manufacturers publish the names, but they don't advertise them on their home pages: usually they are buried deep within their technical pages. If you cant find the information you're looking for, it's worth asking for it. You're quite within your rights to do so.
However you get to the information, once you have it, you've found the pot of gold at the end of the rainbow: safe memory upgrades for your sensitive network devices, no warranties invalidated, very significant savings. And no heartburn.
Wednesday 25 February 2009
Check lead times before you specify
Hold on. Don't order that part. At least, not just yet. The network component you've specified for a particular job may appear to be the right one for the job, but, please, check its availability before setting it in stone and making your project dependent on it.
Spikes and delays
The bad news is that unpredictable spikes in demand and manufacturing delays do occur, pushing component lead times up. The good news, on the other hand, is that the components you've selected may well not be the only ones that can do the job you need doing. Often perfectly acceptable alternatives can be identified. By checking availability at the outset, you can specify components that are readily available, in preference to those which are not, taking into account any effects of the component changes as you go.
It is essential, though, to make such checks at the design stage of your project, before placing your hardware order. Changing project specifications later on is complex, expensive and time consuming. For large companies the problems are compounded as back-tracking through the authorisation process consumes enormous amounts of time.
Indeed, the issue is no respecter of company size. Over the last ten years I have, on more occasions than I care to count, helped out even the largest telcos in the UK as they wrestled with it. Typically their usual supplier had been quoting lead times of six weeks or more – if they could quote a firm lead time at all – for a component that they needed there and then.
Costs up, time wasted
It may be that your design can accept an available alternative to the delayed or unavailable component. However, the alternative will typically be a more feature-rich version of the one you originally specified, and therefore more expensive. You've probably wasted hours trying to track down your preferred part, and now your project costs are up as well.
Some seven years ago I sold some Cisco Port Adapters to the largest telco in the UK at the time. They weren't available in the UK at the time, but I tracked some down in California. The only way I could make even a small margin on the sale was to sell them at global list price: no discount whatsoever. The customer paid it.
I cannot, therefore, over-state the case for investing a small amount of time, up front, in researching the availability of the components you are considering specifying for your project. This can be as straightforward as a quick phone call to your reseller or distributor, to ask if there are any issues with current lead times on your proposed kit list.
It may well save your sanity, as well as your project.
Spikes and delays
The bad news is that unpredictable spikes in demand and manufacturing delays do occur, pushing component lead times up. The good news, on the other hand, is that the components you've selected may well not be the only ones that can do the job you need doing. Often perfectly acceptable alternatives can be identified. By checking availability at the outset, you can specify components that are readily available, in preference to those which are not, taking into account any effects of the component changes as you go.
It is essential, though, to make such checks at the design stage of your project, before placing your hardware order. Changing project specifications later on is complex, expensive and time consuming. For large companies the problems are compounded as back-tracking through the authorisation process consumes enormous amounts of time.
Indeed, the issue is no respecter of company size. Over the last ten years I have, on more occasions than I care to count, helped out even the largest telcos in the UK as they wrestled with it. Typically their usual supplier had been quoting lead times of six weeks or more – if they could quote a firm lead time at all – for a component that they needed there and then.
Costs up, time wasted
It may be that your design can accept an available alternative to the delayed or unavailable component. However, the alternative will typically be a more feature-rich version of the one you originally specified, and therefore more expensive. You've probably wasted hours trying to track down your preferred part, and now your project costs are up as well.
Some seven years ago I sold some Cisco Port Adapters to the largest telco in the UK at the time. They weren't available in the UK at the time, but I tracked some down in California. The only way I could make even a small margin on the sale was to sell them at global list price: no discount whatsoever. The customer paid it.
I cannot, therefore, over-state the case for investing a small amount of time, up front, in researching the availability of the components you are considering specifying for your project. This can be as straightforward as a quick phone call to your reseller or distributor, to ask if there are any issues with current lead times on your proposed kit list.
It may well save your sanity, as well as your project.
Tuesday 20 January 2009
Keep engineering resources in-house
Outsourcing works well in many areas of business activity, but I want to sound a note of caution. In fact, when I hear about companies taking their network management operations out of the safe hands of their experienced and committed in-house engineers and handing it to external providers, my blood runs cold.
Unfortunately, business owners and directors often focus too closely on the savings promised by network management outsourcing, while seriously underestimating the importance and value of reliability in their organisations' IT networks, and the enormous damage that failures can cause.
Network downtime: a company-killer
"I can't get web access and I can't check my emails. I might as well go home." It's a common complaint in today's office, and it illustrates how dependent we have become on our company networks.
Ten years ago we would simply have turned to the phone and fax. For years, though, IT networks have been becoming steadily more important to companies of all types and sizes and for most organisations they are now absolutely key to all day-to-day operations. Downtime can mean service disruption, dissatisfied customers, irritated staff, regulatory compliance failure, and significant financial loss.
Company networks are no longer just important pieces of infrastructure. They have gone far beyond that. Today, a company's network is its central nervous system. If it fails, the company is paralysed. If normal service isn't restored quickly, the company dies. It's that simple, and it's that important.
PR fallout: it's not pretty
In today's online world, news – bad news, in particular - spreads brutally fast. If a major company's network is down for even a few hours, we hear about it from bloggers, or even the media, almost immediately.
The longer the interruption to normal service persists, the more widespread is the bad publicity, the more serious the PR damage and the more customers who will simply go elsewhere, often permanently. Even if they don't result directly in company failure, network outages do serious damage to company reputations.
Downtime is not the minor inconvenience it once was. It's not just a matter of not being able to check emails or browse websites. It's a potential company-killer.
Protecting your network
In-house engineers' jobs depend directly on their ability and commitment to maintain the company network in good order. That accountability for problems, poor performance and downtime is highly motivating, keeping the company's interests at the top of their priority list: they will care far more deeply about the company's success than any external specialist ever will. As a result they are the company's best option to get the job done right, keep the network working effectively and efficiently, and rectify faults that do arise in short order.
Conversely, handing over responsibility for the ongoing management and maintenance of the network to unknown and untested individuals with no direct interest in the company's well-being significantly increases the danger of network downtime, and the damage it inevitably causes.
Think before you outsource
For businesses with in-house engineering resources, the temptation to cut costs by outsourcing should therefore generally be avoided.
That's not to say that outsourcing is never the right option. Many businesses don't have the resources at their disposal to effectively manage and maintain their networks. For such businesses, outsourced network management can be invaluable. But even in such cases, the business should employ at least a few engineers to work with the external team, acting as its eyes and ears, ensuring that the goals of the outsourcing exercise are actually delivered.
Businesses with skilled engineers on the payroll, though, have an enormous advantage. Those engineers are the best protection possible against costly, damaging, and potentially fatal network failures.
Think – very carefully – before outsourcing.
Outsourcing works well in many areas of business activity, but I want to sound a note of caution. In fact, when I hear about companies taking their network management operations out of the safe hands of their experienced and committed in-house engineers and handing it to external providers, my blood runs cold.
Unfortunately, business owners and directors often focus too closely on the savings promised by network management outsourcing, while seriously underestimating the importance and value of reliability in their organisations' IT networks, and the enormous damage that failures can cause.
Network downtime: a company-killer
"I can't get web access and I can't check my emails. I might as well go home." It's a common complaint in today's office, and it illustrates how dependent we have become on our company networks.
Ten years ago we would simply have turned to the phone and fax. For years, though, IT networks have been becoming steadily more important to companies of all types and sizes and for most organisations they are now absolutely key to all day-to-day operations. Downtime can mean service disruption, dissatisfied customers, irritated staff, regulatory compliance failure, and significant financial loss.
Company networks are no longer just important pieces of infrastructure. They have gone far beyond that. Today, a company's network is its central nervous system. If it fails, the company is paralysed. If normal service isn't restored quickly, the company dies. It's that simple, and it's that important.
PR fallout: it's not pretty
In today's online world, news – bad news, in particular - spreads brutally fast. If a major company's network is down for even a few hours, we hear about it from bloggers, or even the media, almost immediately.
The longer the interruption to normal service persists, the more widespread is the bad publicity, the more serious the PR damage and the more customers who will simply go elsewhere, often permanently. Even if they don't result directly in company failure, network outages do serious damage to company reputations.
Downtime is not the minor inconvenience it once was. It's not just a matter of not being able to check emails or browse websites. It's a potential company-killer.
Protecting your network
In-house engineers' jobs depend directly on their ability and commitment to maintain the company network in good order. That accountability for problems, poor performance and downtime is highly motivating, keeping the company's interests at the top of their priority list: they will care far more deeply about the company's success than any external specialist ever will. As a result they are the company's best option to get the job done right, keep the network working effectively and efficiently, and rectify faults that do arise in short order.
Conversely, handing over responsibility for the ongoing management and maintenance of the network to unknown and untested individuals with no direct interest in the company's well-being significantly increases the danger of network downtime, and the damage it inevitably causes.
Think before you outsource
For businesses with in-house engineering resources, the temptation to cut costs by outsourcing should therefore generally be avoided.
That's not to say that outsourcing is never the right option. Many businesses don't have the resources at their disposal to effectively manage and maintain their networks. For such businesses, outsourced network management can be invaluable. But even in such cases, the business should employ at least a few engineers to work with the external team, acting as its eyes and ears, ensuring that the goals of the outsourcing exercise are actually delivered.
Businesses with skilled engineers on the payroll, though, have an enormous advantage. Those engineers are the best protection possible against costly, damaging, and potentially fatal network failures.
Think – very carefully – before outsourcing.
Subscribe to:
Posts (Atom)