100% uptime: Systems Engineer

Discussion in 'Employment & Jobs' started by Sparky, Jul 17, 2007.

  1. Sparky
    Highly Decorated Member Award 500 Likes Award

    Sparky Zettabyte Poster Moderator

    10,718
    543
    364
    Based on my studies for the MCSE MS says it is feasible to achieve 95% uptime for your server infrastructure but to get it to 99.9% it takes serious investment. Even in some of the adverts for SQL server in the small print there is a reference to 99.9% being the maximum uptime.

    Despite this while I was hunting around the job sites today there was a company advertising for a Systems Engineer and one of the requirements was to achieve 100% uptime for their server environment. Reading on they only have 3 Windows servers. WTF?

    Has the world gone mad? :blink
     
    Certifications: MSc MCSE MCSA:M MCSA:S MCITP:EA MCTS(x5) MS-900 AZ-900 Security+ Network+ A+
    WIP: Microsoft Certs
  2. r.h.lee

    r.h.lee Gigabyte Poster

    1,011
    52
    105
    Sparky,

    Is it possible that this company is a pure "internet only" company without a brick and mortar infrastructure?
     
    Certifications: MCSE, MCP+I, MCP, CCNA, A+
    WIP: CCDA
  3. Theprof

    Theprof Petabyte Poster

    4,607
    83
    211
    I understand that the less down time the better, but you have invest money on the technology that provides that and even then its not 100% guaranteed.
     
    Certifications: A+ | CCA | CCAA | Network+ | MCDST | MCSA | MCP (270, 271, 272, 290, 291) | MCTS (70-662, 70-663) | MCITP:EMA | VCA-DCV/Cloud/WM | VTSP | VCP5-DT | VCP5-DCV
    WIP: VCAP5-DCA/DCD | EMCCA
  4. hbroomhall

    hbroomhall Petabyte Poster Gold Member

    6,624
    117
    224
    100% uptime is *very* difficult to achieve. It has been done - a company who pitched to us last week claimed this, but the cost was *huge*.

    In the 'top quality' world "5 nines" is a popular target. This takes *serious* effort to deliver.

    Harry.
     
    Certifications: ECDL A+ Network+ i-Net+
    WIP: Server+
  5. BosonMichael
    Honorary Member Highly Decorated Member Award 500 Likes Award

    BosonMichael Yottabyte Poster

    19,183
    500
    414
    If you decide to apply, tell them exactly what it would cost for 100% uptime, and why. If they're smart, they'll:

    1) hire you, and
    2) realize that it's not usually cost-feasible.
     
    Certifications: CISSP, MCSE+I, MCSE: Security, MCSE: Messaging, MCDST, MCDBA, MCTS, OCP, CCNP, CCDP, CCNA Security, CCNA Voice, CNE, SCSA, Security+, Linux+, Server+, Network+, A+
    WIP: Just about everything!
  6. Fergal1982

    Fergal1982 Petabyte Poster

    4,196
    172
    211
    It is physically impossible to achieve 100% uptime. It just cant be done. Most Website hosts advertise uptime similar to 99.9% uptime. They achieve this with multiple redundancies, power generators, routing, off-site servers with copies, designed to take over if the original is unavailable.

    However, stating you have a 100% uptime is just stupid. It is feasable (However improbable it may be after your precautions) that some sequence of events bring your availability down, killing your 100% uptime.

    To be honest I likely wouldnt apply for this job. Stating that you are responsible for ensuring 100% uptime sounds like scapegoat syndrome to me.
     
    Certifications: ITIL Foundation; MCTS: Visual Studio Team Foundation Server 2010, Administration
    WIP: None at present
  7. BrizoH

    BrizoH Byte Poster

    243
    6
    25
    Tier 4 datacentres costing tens of millions will only commit to 99.995% uptime - whoever wrote the ad for this company clearly doesn't have a clue what it would take to achieve
     
    Certifications: CCNA, CCNA Security
    WIP: CCNP
  8. fortch

    fortch Kilobyte Poster

    408
    21
    35
    The ultimate goal is 6 9's -- 99.9999, which amounts to a grand total of somewhere around 32 seconds. Sorry, but a reboot takes longer than that most times on a Windows box running most services, not to mention a DC. 5 9's is a bit over 5 minutes, and even that is rough.

    To be fair, Windows Datacenter (which claims these figures) *only* runs on a complete packaged hardware solution, fully tested and compliant from the big OEM's.
     
    Certifications: A+,Net+,Sec+,MCSA:Sec,MCSE:Sec,mASE
  9. Sparky
    Highly Decorated Member Award 500 Likes Award

    Sparky Zettabyte Poster Moderator

    10,718
    543
    364
    Exactly! :biggrin
     
    Certifications: MSc MCSE MCSA:M MCSA:S MCITP:EA MCTS(x5) MS-900 AZ-900 Security+ Network+ A+
    WIP: Microsoft Certs
  10. Phoenix
    Honorary Member

    Phoenix 53656e696f7220 4d6f64

    5,749
    200
    246
    Actually
    thats not true anymore
    Datacenter can be purchased through specific license agreements now

    I dont think it claims those numbers any more though :)

    I think its worth actually comprehending the differences in cost between 4, 5 and 6 nines uptime
    the cost and logistic implications are quite astronomical in reality
    and in the scheme of things were talking seconds and minutes of downtime a YEAR at those figures
    very VERY few systems require such measures, were talking about things like control systems in nuclear reactors, not even core systems in finance houses can tout that sort of uptime, and they dont actually need it either!

    personally its VERY unrealistic to assume 100% uptime for any system in the corporate world, its just not required for 99.9999% of corporate entities!
     
    Certifications: MCSE, MCITP, VCP
    WIP: > 0
  11. zebulebu

    zebulebu Terabyte Poster

    3,748
    330
    187
    Utter tosh. Whoever wrote that job spec doesn't have a Danny what they're going on about.

    It is ABSOLUTELY impossible to guarantee 100% uptime. ABSOLUTELY impossible. By the same rationale, it is ABSOLUTELY impossible to guarantee five nines, or six nines or whatever the figure is. You can work toward eliminating all downtime - and have your SLA set at 99.999%, but unless you have a completely triangluated infrastructure - with all things like power, comms, access and everything else that goes along with securing and maintaining that infrastrcuture, spread over an area large enough to not be affected by any major disaster (natural or man-made) then you can't quote uptime figures at that range.

    The 'five nines' figure is standard for a lot of datacentres now and SAN, Virtualisation and all sorts of other fantastic technologies make maintaining this a pretty solid probability.

    I don't think I'd want to work for a company that was foolish enough to tempt fate by saying they had a '100% uptime' SLA - you're just asking for some idiot road crew to go through a massive cable run somewhere, or for a fire to break out in a metropolitan cable duct (as happened in the telco fire in Manchester a couple of years ago). Add to this the fact that there are more and more gangs of thieves nicking cable from places like live railway lines to sell on to China and it doesn't take a genius to see that this sort of thing is likely to have a drastic effect on availability promises in the next few years.

    And I concur with the earlier post - it sounds like they're looking for a scapegoat to blame when something goes wrong with their ESX boxes or their SAN...
     
    Certifications: A few
    WIP: None - f*** 'em

Share This Page

Loading...
  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.