Follow Slashdot stories on Twitter


Forgot your password?
Bug Upgrades Hardware

What To Do When an Advised BIOS Upgrade Is Bad? 467

Bomarc writes "Twice now I've been advised to 'flash the BIOS to the latest,' once by a (major) hard drive controller maker (RAID); once by an OEM (who listed the update as 'critical,' and has removed older versions of the BIOS). Both times, the update has bricked an expensive piece of equipment. Both times, the response after the failed flash was 'It's not our problem, it's out of warranty.' Given that they recommended / advised that the unit be upgraded, shouldn't they shoulder the responsibility of BIOS upgrade failure? Also, if their design had sockets rather than soldering on parts, one could R/R the faulty part (BIOS chip), rather than going to eBay and praying. Am I the only one that has experienced this type of problem? Have you been advised to upgrade a BIOS (firmware); and the upgrade bricked the part or system? If so, what did you do? Should I name the companies?"
This discussion has been archived. No new comments can be posted.

What To Do When an Advised BIOS Upgrade Is Bad?

Comments Filter:
  • by Johnny Loves Linux ( 1147635 ) on Sunday February 10, 2013 @02:47PM (#42851511)
    It's been almost 4 years since I built my last box. I'm planning on building another desktop this summer and would like to know who to avoid as I'm intending to purchase a motherboot that's supported by coreboot so I don't have to deal with UEFI. If there's a motherboard vendor doing evil stuff and they're listed I would like to avoid them if I can. Here's the link for supported motherboards: []
  • Re:Yes (Score:5, Informative)

    by Bomarc ( 306716 ) on Sunday February 10, 2013 @02:54PM (#42851559) Homepage
    Hard drive RAID controller: by LSI []

    System: Dell [] PE 1950; critical update for the BMC controller.

    ... BTW: EMS firmware upgrade for the BSM [] V 2.50 bricked two motherboards. The motherboard for system #1 *may* have had a faulty BMC, however system #2 was working perfectly.

  • Re:Yes (Score:5, Informative)

    by Bomarc ( 306716 ) on Sunday February 10, 2013 @03:03PM (#42851623) Homepage
    LSI: I ran into problems with later hard drives being recognized. I was advised to flash the BIOS. It bricked the controller. After the advice of "buy another"; I removed all of the (in service) LSI controllers, as the risk of data loss (and the need for later hard drive compatibility) forced me to remove them.
  • Car (Score:2, Informative)

    by zmooc ( 33175 ) <.ten.coomz. .ta. .coomz.> on Sunday February 10, 2013 @03:12PM (#42851683) Homepage

    I have a 1996 car. Sometimes things break or I want to alter something which requires modifications. Often, I turn to the service manual. If, while following the directions in it, something goes horrible wrong, I wouldn't even consider holding the manufacturer of the car responsible. That's what you should do too: deal with it. Take your loss, don't use outdated equipment or have it serviced/modified/upgraded by professionals that have insurance that covers these kinds of risks, which happens to be the single most important thing that makes professionals professionals.

  • Re:Yes (Score:2, Informative)

    by PsychoSlashDot ( 207849 ) on Sunday February 10, 2013 @03:16PM (#42851733)
    So, you've got one or more servers which have lapsed warranty. You applied BIOS updates and bricked your controllers.

    I don't mean to be a knob but I think the fault doesn't particularly lie with the vendor. Unless the update bricks most or all cards out there, it's more likely your config or procedure resulted in this. The bottom line is that you're running a non-warrantied configuration and something something something, resulting in bad. It doesn't matter what the something is, nobody's obligated to support a set of hardware that doesn't have support maintenance in place on it. I absolutely cringe for you... your situation totally sucks, but even if the update was named "OMGWTF PONIES! CLICK HERE!", you still did a maintenance function on a machine while lacking the standard support safety net.

    Realistically, even just USING the server is at-your-own-risk. Anything you do beyond "shut it off and replace it" is - sadly - your own circumstance.
  • Re:Yes (Score:5, Informative)

    by Anonymous Coward on Sunday February 10, 2013 @03:17PM (#42851745)

    What's the failure mode on each of {the BMC, the PERC}? I have some experience handling failures of this nature.

    In particular, it's been my experience that on some Dell models of that generation, if you update the BMC firmware without updating the NIC firmware as well, the BMC will fail to be reachable on the network. Fortunately a NIC F/W update fixes this readily enough.

    I wish they told you that.

    [Too lazy to log in.]

  • Ummmm, no (Score:5, Informative)

    by Sycraft-fu ( 314770 ) on Sunday February 10, 2013 @03:34PM (#42851861)

    That is not at all the case. BIOS/firmware/driver updates/upgrades can potentially do four things for a working system:

    1) Add new features. Many products get new features as their life goes on. My desktop board, an Intel, has gotten a number of new BIOS features during its life. When you update the code that runs something, no surprise that code can add features.

    2) Improve performance. Sometimes, a faster/more efficient way of doing something is discovered. It takes an update to make that happen. I remember a big one back in the day with 3com switches. A firmware update provided a major improvement in through put and CPU usage.

    3) Fix a bug that you haven't hit yet, but could. This is why you'll see updates tagged as urgent. Just because you never hit a bug that got discovered, doesn't mean the bug isn't there. So you want to get it fixed, BEFORE you hit it. There have been firmware updates that fixed some nasty ones, like data corruption with SSDs. Some people never got hit, but that doesn't mean the update wasn't a good idea.

    4) Security issues. Same deal as with the bugs, just a different kind of bug. If a security issue is discovered, it'll take a patch to fix it and the system will be working before the patch.

    The "Don't fix it if it ain't broke," really is not a valid ideology for systems administration.

  • Re:Yes (Score:4, Informative)

    by Bomarc ( 306716 ) on Sunday February 10, 2013 @03:39PM (#42851913) Homepage
    "Lapsed" warranty is not fair. They (Dell) won't let me renew it. Up to the (Urgent: Recommended) flash upgrade, the systems worked fine, and were being re-purposed.

    Also, do you think that they (the manufacturer) is going to say (or admit) or have a warning that says: "66% of the people that applied this critical update bricked their system" ??

    As for "OMGWTF PONIES! CLICK HERE"; there is a radical difference between "critical by manufacturer" and "ponies"

  • Re:Yes (Score:2, Informative)

    by Anonymous Coward on Sunday February 10, 2013 @04:10PM (#42852195)

    There is actually. LSI has a dualopoly with PMC Sierra, and PMC owns the adaptec brand. So if you buy an adaptec RAID controller then you are buying a PMC chipset.

  • Re:Yes (Score:4, Informative)

    by maestroX ( 1061960 ) on Sunday February 10, 2013 @04:23PM (#42852309)

    Ever since 3ware was assimilated by LSI, there aren't many viable alternatives to them, if you must have hardware RAID.


  • Re:Yes (Score:5, Informative)

    by MikeBabcock ( 65886 ) <> on Sunday February 10, 2013 @04:33PM (#42852359) Homepage Journal

    This is one of the many reasons I order Dell servers.

    One of the others being that their next-business-day 5 yr warranty really means next-business-day.

  • Re:Yes (Score:4, Informative)

    by Anonymous Coward on Sunday February 10, 2013 @04:44PM (#42852475)

    Or, you know... Use RAID how it's intended: To guard against disk failure. It's never a replacement for backups.

    RAID is for availability of the system, not for keeping your data safe.

  • Re:Yes (Score:5, Informative)

    by AmiMoJo ( 196126 ) * <mojo@world3.nBLUEet minus berry> on Sunday February 10, 2013 @05:19PM (#42852845) Homepage Journal

    As usual Slashdot posters don't mention where they are, but in the UK you would still be able to get the controller fixed or partially refunded (plus possibly costs incurred due to having to switch brand) thanks to the Sale of Goods Act. Doesn't your country have any consumer protection from douchbag vendors?

  • You are thinking of Gigabyte motherboards. Dual-BIOS has been standard for over a decade on those.

  • Re:Yes (Score:4, Informative)

    by Smallpond ( 221300 ) on Sunday February 10, 2013 @06:31PM (#42853433) Homepage Journal

    The only thing you should need the BIOS flash for is to boot from a RAID. I was assuming if he bricked the controller then it is also the RAID firmware, since otherwise you can boot from a rescue disk and run lsiutil to rewrite the BIOS flash.

    The problem with skipping intermediate versions is that the on-disk format can change (more new stuff in metadata, for example). Each firmware rev only knows its own version and how to update from the previous version.

  • by GigaplexNZ ( 1233886 ) on Sunday February 10, 2013 @10:00PM (#42854845)

    I'm intending to purchase a motherboot that's supported by coreboot so I don't have to deal with UEFI

    Why? What's wrong with UEFI that you need to replace it with coreboot (which just so happens to have a UEFI payload [])

  • Re:Yes (Score:3, Informative)

    by anyanka ( 1953414 ) on Monday February 11, 2013 @12:15AM (#42855579)

    Their support might be inconsistent between regions (or time). I've had the opposite experience; disks replaced based on SMART reporting a imminent failure, big SCSI disks replaced next day and so on. I've been less impressed with HP, and SGI (several weeks to deliver a disk for a relatively new system). But then again, this was a few years back, and in Norway (Dell support subcontracted to a local provider), so your mileage may vary. These days I have less advanced hardware and do repairs myself, with impeccable same-day service. ;)

You are in a maze of little twisting passages, all different.