Slashdot Mirror
Ask Slashdot: Do You Trust When a Vendor Tells You To Buy New Parts?
Nerval's Lobster writes "Roughly 85 percent of IT managers polled by Forrester said they would hold onto networking infrastructure longer, but vendors retire products prematurely in an effort to force customers to upgrade. In a response that may seem familiar to anyone who's ever been pressured into buying a maintenance contract—either by an enterprise vendor or a major electronics retailer—over 80 percent of the 304 respondents said they don't like the misrepresented cost savings, new fees, and inflexible pricing models—but buy the products anyway. One of the survey's interesting points is that IT decision makers aren't willing to contradict the vendor. The uncertainty seems to come from the fact that the vendor may in fact be right—and a customer who contradicts what they're saying may end up shouldering the blame if the equipment goes south. It's the 'you never got fired for buying IBM' argument, applied to the networking space. The problem, of course, is that the vendor often works for its own agenda. Do you upgrade when the vendor (or reseller) suggests you do so? Or do you stick to your own way of doing things?"
And, let's face it, whose money you're spending.
Nearly all HP kit has it even a lot of Cisco kit does (though they make you jump through hoops to use it). Buy good kit I've replaced cisco 6500's bits over the years that were bought in the 90's and just got tech refreshes not bad taking a 10/100 with a few gig ports to 10ge over 14 years.
No sir I dont like it.
percent off list, and since you're looking at a new rollout anyways, do they still think you need new hardware even if it's a competitor's?
If they say yes, you can probably believe. And might save a few quid on the rollout to boot :)
Do You Trust When a Vendor Tells You To Buy New Parts?
Yes, if its a video card. Buying a low end video card (US$120-140) every two or three years seems to improve the end user experience nicely, **iff** we are talking about a system used for gaming.
On second thought I guess I am not really trusting the vendor since they are telling me to buy the US$500 video card.
I'm all for upgrading, when the vendor can show specific reasons why it benefits me.
Here's a related question. Do you trust when a car manufacturer tells you to buy new parts?
Specifically, the maintenance schedule in the owner's packet that comes with a new car. For example, at 60,000 miles:
1) Replace engine coolant
2) Replace HEV inverter coolant
3) Replace manual transmission oil
4) Replace automatic transmission/CVT/eCVT fluid
5) Replace differential oil
6) Replace engine drive belts
7) Replace radiator cap
8) Replace transfer case oil
Are all these necessary, or is the dealer trying to squeeze more money from the owner? I've heard various mechanics coming down on both sides of this question. Does the differential oil really need periodic replacing? Do you need new drive belts if there's no visible damage?
(Also: Do you replace the engine oil and filter every 2000 miles, or is this just another way to squeeze money from the consumer?)
Do you upgrade when the vendor (or reseller) suggests you do so?
The answer is no, they don't. Which is exactly what it should be, and I doubt anyone (that isn't a money grubbing whore in management) minds.
If you have equipment that works, why should you upgrade? It works!
The problem lies when the old equipment becomes run down and faulty, they refuse to update to something that DOES work, and demand that we spend manpower and extra money we'll never get back fixing their outdated machines, which 6/10 times isn't even possible and we just end up wasting everyone's time.
HP will extend the warranty on any business class system they sell for a minimum of 5 years beyond the initial 3 year warranty, at the end of 8 years it probably IS more cost effective to replace the system (hell, the HP 3000 series boxes were supported for over a decade after end of sale). For networking I love Cisco chassis based switches, Cat 6500, 4000, and 4500 series switches have all lasted at least a decade.
There are 4 boxes to use in the defense of liberty: soap, ballot, jury, ammo. Use in that order. Starting now.
That's really the best you can do, and in most environments it's not possible anyways, but if you elevate the decision to a boss and don't advise him one way or another then the decision will rest squarely on their shoulders, do be certain to do this in email of course, and pass along all the spam you get from the company as to why you should upgrade, and even then... sometimes you'll get fingered for it anyways, but generally if an email trail isn't enough to keep the blame from falling on you then your time at that company was limited regardless.
TBH the money a company will end up losing upgrading prematurely isn't so significant that it's worth you risking your job for.
Stop golf course meetings and let real IT people in to meetings as well.
generally i keep my options pretty open. infrastructure servers are usually high availability and ordered from Silicon Mechanics or something. theyre cheap, my management enjoys the cost savings, and if one breaks its super simple to just order another as opposed to trying to justify the 'value.' ERP applications or databases will get the Dell/HP Treatment with the $nonferrous_metal level service support and $mm/$dd/$yyyy response SLA because management sees more value in them and theyre generally easier to get upgrades and DR stuff for. Dell for example knows this and actually ships an SAP "break-down" sheet for my manager to get the fuzzies about so he can look good in front of his management, who in turn can tout our 'core relationship with leading technology vendors' to investors and C-levels.
Ive stayed away from Cisco because of the cost, lock-in, and seriously underhanded sales tactics theyve used in the past. Things like firewalls and VPN are nearly exclusively Open Source here just because management cant justify the cost of a laptop for someone, let alone the cost of a token/license/enterprise server. Management gets their nano-yubikey (which they think is incredibly tech-savvy and sexy) and everyone is assigned a fun password from pwqgen.
Good people go to bed earlier.
If you don't want to upgrade every 2-3 years you could always:
- You're a small shop with no money and the equipment is doing business critical work: Carry a spare and possibly arrange in redundant configurations
- You're a small shop with no money and the equipment is doing nothing critical: Possibly carry a spare
- You're a large shop with 'too much' money and the equipment is doing business critical work: Carry spare(s) and arrange in redundant configurations
- You're a large shop with 'too much' money and the equipment is nothing critical: Carry spare
Never buy equipment that can't be vendor/product line swapped unless you're seriously in bed with the vendor and have an iron clad support contract. Best to mix up equipment from time to time just to make sure that your IT people CAN adapt to other vendors if the sh hits the fan.
Bye!
We needed a caddy anyway.
Not if you're these guys.
Proverbs 21:19
Maybe some SOHO/small business networking gear gets retired prematurely, but for larger scale Cisco stuff the end of sale/end of life dates are way beyond when anyone would reasonably want to use the gear.
There is no memory shortage. yes I have heard of XFCE. Go away.
SonicWall (now Dell SonicWall) does this to no end. Perfectly capable hardware but they stop producing firmware and updates, warranty contracts, etc.. prematurely and force you to the next product.. Great products though BTW.. Just the forced upgrade/migration sucks..
We generally lease servers but buy network equipment. All of our network equipment has redundant units and is a pretty static environment, so I don't run out and buy new equipment every time a vendor end of life's something or recommends we replace it. When we have a failure/rollover situation, if the equipment is end of life, we'll upgrade, if not, and it's still under support, we replace the failed device and life goes on.
IT managers upgrade anyway when pressured because "who is going to bail me out of that tight spot when the excrement hits the fan?" The answer is "Not the vendor. They don't support that anymore and I don't have a current support contract". Warranties and support contracts are like buying insurance. You hope you don't have to use it, but you are a moron if you don't have it.
Here's a related question. Do you trust when a car manufacturer tells you to buy new parts?
Specifically, the maintenance schedule in the owner's packet that comes with a new car. For example, at 60,000 miles:
1) Replace engine coolant
2) Replace HEV inverter coolant
3) Replace manual transmission oil
4) Replace automatic transmission/CVT/eCVT fluid
5) Replace differential oil
6) Replace engine drive belts
7) Replace radiator cap
8) Replace transfer case oil
Are all these necessary, or is the dealer trying to squeeze more money from the owner? I've heard various mechanics coming down on both sides of this question. Does the differential oil really need periodic replacing? Do you need new drive belts if there's no visible damage?
(Also: Do you replace the engine oil and filter every 2000 miles, or is this just another way to squeeze money from the consumer?)
You're talking about consumables. What the vendors are doing is the same as a car manufacturer telling you to buy a new car because it's out of date - regardless if it still works or not.
I'm sure most of us have dealt with sales reps over the years, and seen all sorts of claims of bigger/better/faster/cheaper, but they're often unsubstantiated by anything.
We had a scenario with a vendor a while back where functionality we were relying on wasn't going to be in their next version until a year after it was too late for us. (Add on component we'd been using for years.)
So, we basically forced them into extending support since the only reason we couldn't upgrade was because of their inability to deliver functionality we already had.
Then they spent the next year constantly asking us when we would be upgrading, and conveniently trying to forget about the signed contract they'd given us to extend support and telling us we were about to become unsupported.
You need to work with your vendor, but you sure as hell don't need to take what they tell you at face value without something to support it.
At the end of the day, most of the salesmen (because that's what your rep is) are more worried about their commission check than anything else, and will certainly mislead your or pressure you to do something which doesn't really benefit you.
Lost at C:>. Found at C.
All of our vendors are made aware of our purchasing policy.
We do not purchase products from the first vendor who tells us we have to have them. If we do need something, we purchase it from another vendor.
For example, a vendor comes in and tells us we need a new router. We do the analysis in house and if we determine we do in fact need a new router, we buy it from anyone but the guy who told us we need a new one.
Cuts down on the bullshit.
You continue to use the product if the following requirements are all met:
A) The hardware continues to meet the performance SLA.
B) You have access to replacement parts in a way that meets your availability SLA.
C) You continue to receive security patches.
Everything else is circumstantial.
Well, we're actually an integrator. We generally recommend new gear for two reasons: Age & insufficiency. We understand that the gear we sell is the lifeblood of many of the companies that use it. We understand that even if a 5 year old raid is sufficient to the task that it will soon begin to start failing. These are under support and the parts will be replaced quickly, but the potential for a catastrophic failure rises dramatically. Additionally, if we wait until year 6 to start pressing for replacement then it will likely be well into year 7 before it's replaced. In the case of sufficiency, often with newer hardware (raids/SANs & tape libraries specifically) there are substantial performance improvements to tbe had. When a customer grows because of the products we sell they will often come back at some later date and complain of performance issues. We try to spec gear correctly so that there is headroom, but often the customer will not have the budget for the headroom. So if you buy at the low-end you don't have that flexibility moving forward. In other words not all vendors are just trying to squeeze you for more money. Well thought out purchases will likely be useful longer if you plan correcty, even if you're paying more. Additionally, respected integrators with good references can often help you get exactly what you need even if you have no clue. Yes, that's a self-serving plug for my industry, but assuming you get actual customer references and speak with the customer then you should be much better off.
Never ask a tire salesman whether you need new tires.
What if they don't want to share their cut of the bribe with the IT people?
All of these Ask Slashdot questions appear to come from a Computer Science freshman class.
Before you buy a piece of hardware, you find out what the "lifetime" is of a piece of hardware and how long it is going to be supported by the vendor. For example, my last job we bought Dell servers, so we investigated that we could get support contracts from Dell for 5 years after purchase. So, we used servers in production for 3 years, and then after that, we would rotate them to another production level function that did not need the latest level of CPU or memory, and then at the end of 5 years, they were rotated to non-production. They stayed in non-prod until they died and we could not easily get parts for the server.
If you depend on something for production, you have to follow the vendor lifetime guidelines. If the vendor does not make it clear what the lifetime is, then don't buy from that vendor.
Linux O Muerte!
"and a customer who contradicts what they're saying may end up shouldering the blame if the equipment goes south. It's the 'you never got fired for buying IBM' argument, applied to the networking space."
Since when has "management based on fear" ever been a good way to run a department?
If you are really so afraid that you will buy expensive equipment that is probably unnecessary in order to keep your job, then either:
(A) you should lose your job, you coward. Or
(B) you are in a toxic workplace and need to find another one right away.
Why are "IT decision makers" listening to anyone outside their organization when it comes to actual decisions?
I'm sorry, but they have many people in their organizations who have informed opinions on equipment - such as the people who have to work with them. Things like, "These NetApps are shit, let's go with someone else" or "we need new switches, these are dropping packets and are totally fabric saturated". Employees tell their bosses this stuff all the time; they know it amongst themselves as well.
What's more, it takes what, 15 minutes to get a feel for how bad a product is online before purchasing it - 30-60 if you don't really know what you're looking for or aren't too familiar with the technology?
Why are purchasing deciders making decisions in a vacuum when there is more than enough information available?
~/ssh slashdot.org ssh: connect to host slashdot.org port 22: too many beers
It's like the lifetime warranty for TomTom GPSs : the warranty applies for as long as the device is functional; once the product stops functioning, the lifetime warranty no longer applies. WTF?
HP Procurve devices break so often and so frequently that we automatically get upgrades. If we try to send back a shelved device, we just get the next model up.
If it stops functioning, it's lifetime is over. Duh. :)
We recently decommissioned a perfectly good Sonicwall CDP-6080 with 4TB of backup storage. We were only using about half of the device's available storage, and it was plenty fast for our needs, yet Sonicwall/Dell would not renew our service contract for the device. We were simply told to buy the replacement model in the lineup.
At the time (4 years ago), it was one of the few backup appliances that could handle AD/Exchange/SQL/Linux off-site backup and manual external archiving to disk. It was expensive but it fit the bill perfectly, and it wasn't Backup Exec...
Fast forward a few years, and we've put our web server, RDP/SQL farm, and email in the cloud. We figured we would simply keep the CDP appliance as an AD/file server backup device. Since Dell/Sonicwall refused our support renewal, we put our backup in the cloud.
I suspect as cloud services become more popular, hardware vendors will have less leverage in pushing unnecessary hardware upgrades on their clients.
Most times it is all about budget. If you can purchase a fully redundant infrastructure environment, you can suffer end-of-life failures and replace those devices with newer equipment, even out of warranty. You also get the benefit of having backup in the event of an infrastructure failure. Ideally, we would all have a fully redundant infrastructure to start with and then periodically rip it all out and replace it with newer/modern equipment after EOL/EOS...the problem is that, in my experience, the new products completely replace the old products within a few years, thus making the choice for you. The best approach you can have IMO, is a staggered one that allows you to operate and use the equipment you have already while improving in critical areas such as your core switching/routing and then staging replacements of less critical resources like edge switches. It goes like this:
Day 1: Purchase initial gear for company and cold standby spares for edge devices if possible
3 years later: Purchase redundant distribution and core infrastructure devices with newer technology as well as approx. 1 edge switch with newer technology for every 5 edge devices. Migrate all critical routing and switching tasks to new infrastructure as the primary path while still allowing failback to the older, redundant architecture (even at a loss of speed...your goal is uptime). Also migrate critical resources at the edge to the newer switches while keeping the older switches as cold standby spares.
3 years later: Completely replace all EOL/EOS distribution and core infrastructure with newer technology and sell the older equipment on eBay. Also purchase edge switch replacements for all devices older than 4 years if possible. The newer switches should be kept as cold standby spares.
The reasoning behind a complete edge switch replacement every 6 years is due to technology improvements (i.e. FastE -> GbE -> 10GbE -> 40GbE -> etc.) This may not be important for all environments and it may not be possible as every 6 years you have a higher expense in edge replacements than any time during the interim. A way around this is to purchase spares or replacements yearly as part of a Capitol Expenditure budget rather than a projects budget requiring approvals.
Ultimately you end up with redundant core infrastructure that is never older than 6 years and with active core infrastructure that is never older than 3 years. Also, you either stagger replacements of edge switches as well as having cold spares or you replace all edge switches every 6 years and still have cold spares in the event of a device failure. You end up keeping up with technology (albeit maybe a year or so behind the 'latest and greatest' depending on budget), you have redundancy in the event that a device is EOL/EOS and cannot be replaced, and you have periodic refreshes of the redundant architecture that keeps your critical services running.
Never, ever, ever, listen to the vendor. Remember that when you're talking to them, you're talking to their salesmen. Not their technical lead, not their developers, their salesmen. So many people forget this when they go into these contract negotiations. These people are going to say whatever they can to get you to buy, and they rarely have any idea of what they're talking about.
The details of how long you should expect their product to serve your needs should be explicitly detailed in your contract before you buy. "this equipment/software will work for X years" If they come to you before X years is up and tell you that you need new, then you need to settle up on that old contract that they've just invalidated. If you're beyond what was contractually stipulated, well that's a management call. How important is the application of the product? If it fails, what's the impact? Does it make it harder for people to do their jobs or will you have 2000 people sitting idle until you have it fixed? Also, never sign "their" contract. We have a "contracts" department that writes all of ours, and often have vendors try to force their own contract down our throats. But, we have a company policy that stipulates we NEVER sign with the vendor. The vendor always signs with us. It's our contract, in our words, written by our legal team. If they don't like it there's nothing we can do about it, they cannot do business with us.
But keep in mind, negotiating a new contract when your completely screwed and need it now now now, is not the situation you want to be in. On the same token, if the vendors coming to you when your old systems running great you have a lot of time to haggle and little incentive to hurry the process. I think weather you upgrade is entirely dependent on what the vendor is offering. If it's a good deal and you're in a position to argue for an even better deal then why not? But if everything's running great and the vendors coming to you with some Y2K nonsense and you need to upgrade now at 2x the price of your last upgrade then tell them to shove it.
One of the survey's interesting points is that IT decision makers aren't willing to contradict the vendor.
Then they're shitty IT guys.
Seriously - if some salesman is going to tell you how and what to do with your job, if they can't present you with a convincing argument you tell them to STFU and move the fuck on. Of course, some industries really do need to be on the bleeding edge, but the vast majority do not. The VENDOR should not be making decisions about your purchases...that's like "Buyer Beware 101" right there.
"We'll be happy to fix it unless it's broken."
It could be worse. The definition of "lifetime" may be "lifetime of the registered owner", in which case Cisco would dispatch a ninja assassination squad as soon as you open a warranty ticket.
Welcome to the Panopticon. Used to be a prison, now it's your home.
...But ultimately question it's source and relevance prior to taking it.
In one of our sites we have Five iSCSI storage devices: two HP MSA's two Thecus and an old Qnap for 'warm' archive.
The HP's cost a lot, the disks alone for one of them would pay for the 3 non HP devices. They are the only two devices with a purchased warranty, which cost alot.
Between the two HP's we have had a replacement part delivered and fitted every 3 months for 6 years. In 5 years the other 3 have had 1 disk pop.
I didn't want to buy the HP's or their warranty. But had to purchase something 'enterprise' with a 4 hours SLA. Every time an HP part shows up I hear "Oooh good service; glad we bought that!" - not a peep regarding the other stuff which is just getting on with it.
My VW Golf gets the manufacturers once a year recommended service. Preventative maintenance - think of it as insurance for your car's engine and drive train.
Doesn't seem excessive to me.
It gripped her hand gently. 'Regret is for humans,' it said.
vendors retire products prematurely in an effort to force customers to upgrade
Really..
Never trust a salesman when he wants to make you buy something. He gets bonus based on how much you buy, which means he has an incentive to fool you. A good test is suggesting that if you replace gear now, you could switch to another supplier.
The simple solution is one that has clear benefits: leasing.
1) You refresh your hardware every three years.
2) You don't end up with eight-year-old legacy systems that threaten to die at any time.
3) You're forced to keep your software "fresh."
4) Each new generation uses far less energy for far more computing power.
Kriston
And BFG's lifetime warranty means the lifetime of the company.
We did a cost-benefit analysis and it turned out to be a lot cheaper to have him killed.
Sleep your way to a whiter smile...date a dentist!
if they had good parts I would not be needing new ones as quickly, and they would not cost a left nut for best buy gear in a metal box
machine vendors generally suck, now that will be 149$ for a gasket please
...he said to buy new pants.
I see even classic Slashdot is now pretty much unusable on dial up anymore.