back to article Cisco shipped UCS servers with rotten RAID settings

If you've been wondering about the server performance in your Cisco Business Edition 6000/7000 telephony system, wonder no more: The Borg has issued a field notice that the system shipped with misconfigured RAID. The Cisco field notice advises sysadmins that the correct settings for the kit are as follows: Read Ahead Policy …

  1. Warm Braw

    Engineers forgot to update the installation script

    And presumably nobody from QA attemped an installation before shipping the product?

    If there is anyone from QA in the wake of the recent lay-offs.

    1. Anonymous Coward
      Anonymous Coward

      Re: Engineers forgot to update the installation script

      We did away with our QA team, were useless.

      You need balled guyz in your QA teams that don't gobble the "that is by design" mantra (default dev response to issues), too many, however, are happy with that sentence regardless of the problem.

      1. Anonymous Coward
        Anonymous Coward

        Re: Engineers forgot to update the installation script

        "We did away with our QA team, were useless."

        I'm chuckling since this occurred out of my old team, in a code area I used to update. I was also one of the SW Devs who DIDN'T use the "by design" argument (well... very often anyway) and didn't yell at the QA folks for finding bugs in my crap... I was fine with QA finding bugs, because customers finding bugs was much more unpleasant. :)

    2. Velv
      Facepalm

      Re: Engineers forgot to update the installation script

      Why employ an expensive QA team when the customer will test it for you.

      "It's not working? Oh, yeah, my bad. Here Mr Customer, just change these settings and reinstall and everything will be fine"

      Perhaps if Cisco was required to send an engineer to every installation to fix it they might take a different attitude to QA...

      1. Medixstiff

        Re: Engineers forgot to update the installation script

        "Why employ an expensive QA team when the customer will test it for you."

        That's what Apple do and it's worked for them so far, after all, problems like Antennagate were just user error right?

    3. Captain Scarlet
      Paris Hilton

      Re: Engineers forgot to update the installation script

      Would QA think that it was wrong though to take that long?

      It might have been because it was tested and whilst installing not throwing up "ARGH WE ARE DOOMED press the any key"?

  2. Justin Pasher

    "Standard" strip size?

    Does anyone with one of these devices know what the strip size was (erroneously) set to at the factory?

    In all of the RAID controllers I've worked with over the past decade (granted most were Dell/LSI branded), the default strip size has been 128KB, and tweaking the value has never resulted in a 6x to 7x performance difference for my benchmarks. But considering I don't even know the typical use case of these Cisco devices, it could be an apples to oranges comparison.

  3. Anonymous Coward
    Anonymous Coward

    It's always the caching

    My guess is that the write-back caching makes the biggest difference of these settings. If it was set to write through then installing something (lots of small writes) is going to incur a horrible latency penalty

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Other stories you might like