Will virtualization and consolidation change the way IT buys hardware?

820 pts.
Tags:
Open IT Forum
Server virtualization
Virtualization
Virtualization Hardware
Over the last few years I have had the opportunity to visit a number of large IT server pools (5000 or servers or more) and I was struck by how many different vendors' hardware was represented in each of those pools of servers, it is though the name of the vendor carried no weight in deciding whose hardware to buy. One IT organization had servers from five different vendors and a total of 115 different models. Is there anything about either virtualization or consolidation that will encourage IT management to commit to one vendor's hardware?

Answer Wiki

Thanks. We'll let you know when a new response is added.

I don’t see this making companies standardize on a single vendor or model. I still assume that companies will buy what ever servers that they can get for the best price.

================================================

Just realize that those systems are bought during different times. And with new IT management, they decide on a different vendor platform due to their experience. It’s a merry go around with incoming decision makers to switch vendors midstream. Normally you would want to stick to a single vendor for most part but in certain situation, certain application vendors only certify their products to work on specific hardware.

Discuss This Question: 7  Replies

 
There was an error processing your information. Please try again later.
Thanks. We'll let you know when a new response is added.
Send me notifications when members answer or reply to this question.

REGISTER or login:

Forgot Password?
By submitting you agree to receive email from TechTarget and its partners. If you reside outside of the United States, you consent to having your personal data transferred to and processed in the United States. Privacy
  • Jim4522
    Mrdenny, if you knew that one vendor's server was twice as reliable as the next most reliable server being offered to you and considering that either through virtualization or consolidation you are increasing the amount of work that the hardware will be doing for you, would that cause you to standardize on a single vendor's hardware? For instance if you knew vendor A's hardware caused a disruptive failure on average every 144 "server months" and vendor B's hardware caused a disruptive failure on average every 66 months would that sway you decision. A disruptive failure is one that causes you to lose the hardware until it is fixed, or lose the hardware when it is fixed)? Jim4522
    820 pointsBadges:
    report
  • Denny Cherry
    I'm not sure if that would or not. The server hardware is redundant between nodes in the VM environment so in the event of a hardware failure, everything would still be online.
    66,360 pointsBadges:
    report
  • Jim4522
    Mrdenny, I am sure it is because I don't understand VM as well as you do, but I don't understand your answer to my question. I asked if one server platform was twice as reliable (produced less disruptive events) as another platform wouldn't that sway your decision as to which server platform to virtualize or consolidate on to, your answer was "I’m not sure if that would or not. The server hardware is redundant between nodes in the VM environment so in the event of a hardware failure, everything would still be online." Could you elaborate on your answer? What do you mean by "The server hardware is redundant between nodes" and what do you mean by "so in the event of a hardware failure everything would still be online"
    820 pointsBadges:
    report
  • Denny Cherry
    When setting up a virtual environment (using VMware ESX or vSphere at least) and when you have a SAN backend (which all large deployments do) you build a cluster of servers. This cluster of physical servers can host any virtual machine in the environment at any time so long as there is enough RAM available on the host to fire up the virtual machine. In the event of a hardware failure of the host server, the VMware DRS software will automatically restart any virtual machines which are now offline on other nodes of the cluster with 60 seconds or so. If VMware HT is being used for a virtual machine (vSphere 4.0 and up only) then the virtual machine won't suffer any down time at all when the physical host fails. While it would be nice to have the hardware never fail, virtual environments are (or at least should be) designed to survive a physical failure with little or no impact to the virtual machines other than having those VMs be restarted on another physical host automatically. With a cost difference between machines, and with the auto-recovery which is built into the products having a machine which will fail every 66 months compared with one that fails ever 144 months isn't a big deal. Especially when you consider that most production server hardware is replaced by 36-48 months anyway, then the older hardware becomes dev/qa/dr equiptment.
    66,360 pointsBadges:
    report
  • Jim4522
    Mrdenny, your comment about only using hardware on 36 to 48 months is irrelavent. In a pool of 1,000 servers you have 12,000 server months every year. The difference between a server with an availability rate of 66 months and 144 months is the difference of 181 disruptive failures a year (12,000 divided by 66) and 83 disruptive failure per year (12,000 divided by 144). That means the difference between having the most reliable and the least reliable server is 98 disruptive events per year. To put that in perspective 98 disruptive events comming from your server pool is about 10 times the number of disruptive events that would come from 4 mainframe systems, 20 EMC RAId storage devices and all the centralized teleconumication devices combined in the average IT organization in that same year. In theory having the ability to move workloads from a failing server to some other available server through a resource pool sould allow an IT organization to insolate end users from being impacted by the hardware failure, but you would have to know that the failure that occurred was a hardware failure and not a software failure and I don't think that capability currently exists for the lower end x86 servers. I have just put a two part question out on that issue and taged it under virtualized servers. Thanks I really do appreciate your responding to this conversation. Jim4522
    820 pointsBadges:
    report
  • Denny Cherry
    Even lower end servers should be able to report on hardware failures such as failed RAM and failed hard drives. To reduce the risk of hardware failures some servers allow you to RAID your RAM so that if a stick fails the server will send out an alert, but continue running without issue giving the sysadmin the time to move the VMs off the server and replace the hardware. Combined with RAIDed hard drives to handle disk failures and redundant fans, the odds of a server failure become lower and lower unless there's going to be some sort of major motherboard failure. Combined with Enterprise Class Storage (EMC, IBM, etc) and you should have a solution which can survive most hardware failures and provide an acceptable total up time.
    66,360 pointsBadges:
    report
  • Kevin Beaver
    Price has been and always will be a large factor. Moving forward I think it'll be multiple processor cores, massive RAM and fast network cards.
    17,740 pointsBadges:
    report

Forgot Password

No problem! Submit your e-mail address below. We'll send you an e-mail containing your password.

Your password has been sent to:

To follow this tag...

There was an error processing your information. Please try again later.

REGISTER or login:

Forgot Password?
By submitting you agree to receive email from TechTarget and its partners. If you reside outside of the United States, you consent to having your personal data transferred to and processed in the United States. Privacy

Thanks! We'll email you when relevant content is added and updated.

Following