On 10/29/2009 05:14 AM, Mike Cloaked wrote: > > I have had an unprecented pair of events last night and this morning - > two > machines (both Dell Inspiron 530s desktops) have died completely ( > unbootable) - having been running on f11 very successfully for some time. > I > had no reason to suspect the hardware at all and both were running fine > from > new a couple of years ago - but I am now wondering if this has anything > to > do with updates that came in on 28th October. The new kernel and a number > of > other updates were installed yesterday morning on both machines. Both > were > also using ext4 files systems. Did you reboot after installing the new kernels? I have installed the new updates yesterday as well, but I have not yet reboot to use the new kernels on either my x86_64 laptop, nor an i586 desktop that I use only for testing. Both are still running. Possibly because I have not yet rebooted. Neither have any ext4 filesystems on them. Just a data point for you to compare to.

On 12/31/2009 01:25 PM, Mike Cloaked wrote: > I believe that this relates to > https://bugzilla.redhat.com/show_bug.cgi?id=506171 > > So this bug is not resolved despite the implication in the last post on > the > bz! Obviously if bugs are closed, then there must be less bugs no ;-)

Bugzilla from gabriello.ramirez(a)gmail.com wrote: > > > in the above case, maybe her mistyped her password in one ocassion and > when typed it correctly , the machine got the following bug > https://bugzilla.redhat.com/show_bug.cgi?id=506171 " If password > entered incorrectly in KDM, next successful login causes X server > shutdown and hang " > > maybe you can try to mistype your password and enter correctly with > kernel 2.6.30.8-64 to discard kernel-2.6.30.9-90.fc11.i686.PAE as the > cause > > For your first comment about the response of the system to a mistyped password in kdm I was not aware of - I'll ask if this it was a second login attempt when I talk to her later - and yes i do have kdm as the login manager. The status of the bz you referred to linked to an upstream bug which has not had any reply since the end of August!

On 10/29/2009 02:23 PM, Mike Cloaked wrote: > > In my wife's case she reported that when hitting the enter key to login > under kdm the screen went black, and this was followed by a cursor at top > right and repeated lines containing text, with "nouveau_fifo_free:freeing > fifo 1" > > So this appears to have been a graphics issue concerning the nouveau driver > - however this evening I was able to boot this machine and am running on the > previous kernel for safety. (2.6.30.8-64.fc11.i686.PAE) > in the above case, maybe her mistyped her password in one ocassion and when typed it correctly , the machine got the following bug https://bugzilla.redhat.com/show_bug.cgi?id=506171 " If password entered incorrectly in KDM, next successful login causes X server shutdown and hang " maybe you can try to mistype your password and enter correctly with kernel 2.6.30.8-64 to discard kernel-2.6.30.9-90.fc11.i686.PAE as the cause > The other machine is extinct - I cannot boot at all even to a liveCD so I > can't investigate it. However at the time that failed there were weird > graphics artifacts with multicoloured lines - and it is conceivable that it > could have also been a graphics issue - on boot it puts port00: after the > first message on the screen, and then some stuff on the screen (after the > post check complets) but no boot - I tried partedmagic DVD but it won't boot > at all. I suppose that the disc could have gone bad and over the weekend > will replace the drive and see if that allows a boot and re-install - but I > am out of ideas on that... however the nouveau failure on my wife's machine > worries me.... > if you disconnect the hard drive, maybe you can boot from a livecd, to see if the motherboard and video card are fine. I don't know if the livecd include the memttest maybe the sata cable failed or the sata port where the disk is connected failed, if you have a sata port free you can connect the hard drive in it if using uuids in the partitions Gabriel

waits till you install new software to break. I've seen it happen way too often over the years in our lab at work for it to be a coincidence :-).

I have this long standing theory that hardware deliberately waits till you install new software to break. I've seen it happen way too often over the years in our lab at work for it to be a coincidence :-).

I would like to know if this is a chance event or related to kernel changes, particularly related to ext4?

I am at a total loss as this takes out both operational desktops at home - it is of course possible that it is not connected with any updates - but two machines going out within 12 hours seems a very unfortunate coincidence - last night I was convinced it was an unlucky disk failure - but I have not had time to investigate. However I thought it was worth asking in case anyone else who is running f11, and ran the same updates may have hit the analogous problem - if so I am quite worried as I have several other machines running f11 elsewhere (like on my desk at work!) All are fully up to date f11 - but with different hardware. The two that failed were both Dell 530S less than 2 years old.

I have had an unprecented pair of events last night and this morning - two machines (both Dell Inspiron 530s desktops) have died completely ( unbootable) - having been running on f11 very successfully for some time. I had no reason to suspect the hardware at all and both were running fine from new a couple of years ago - but I am now wondering if this has anything to do with updates that came in on 28th October. The new kernel and a number of other updates were installed yesterday morning on both machines. Both were also using ext4 files systems.

I would like to know if this is a chance event or related to kernel changes, particularly related to ext4? If nobody else has anything remotely similar happen then I will conclude that I am just very very unlucky - but I would like to know if anybody else has had a mysterious total failure of their system in the past 24 hours.

On Sat, 2009-10-31 at 06:24 +0800, Ed Greshko wrote: >> hardly surprising at all...consumer grade SATA hard drives have >> > become > >> unreliable and that is why none of the manufacturer's warranties >> > them > >> beyond a year now. >> >> >> > Really? All of my Seagate drives that I purchase here in Taiwan have > 3 > year warranties. > ---- actually some of the WD and Seagate drives have up to a 5 year warranty but you have to be careful which you buy (I think the Caviar and Barracuda respectively).

The simple truth is as they have increased densities for higher amounts of storage, the error rates have remained constant and they are more prone to failure.

Though I have stopped using RAID-5, this link is pretty sobering... http://www.enterprisestorageforum.com/technology/features/article.php/383... and there are varying grades of SATA drives being sold so there is a need to be cautious.

Ed Greshko wrote: > >> hardly surprising at all...consumer grade SATA hard drives have become >> unreliable and that is why none of the manufacturer's warranties them >> beyond a year now. >> >> >> > Really? All of my Seagate drives that I purchase here in Taiwan have 3 > year warranties. > > > Yeah - well guess which make I just put in? Seagate Barracuda - the old dead one was Samsung HD321KJ - so I can look forward to 3 years of reliable service minimum!

Problem is that when you buy a new machine you don't have control over which make of drive is in it unless you know something I don't!

On Fri, 2009-10-30 at 13:45 -0700, Mike Cloaked wrote: > This evening I removed the HD and installed the brand new hard drive. > The > machine now readily booted to a PartedMagic LiveCD and I was able to > partition the drive to prepare for an f11 clean install which I have > now > just completed. > > So in this case it turned out to be a total failure of the hard drive > - the > machine is less than 2 years old, so a drive failure at this age is > surprising, the more so considering it has always been on a UPS to > protect > the mains supplying it. The new HD is a different make to the > original, and > hopefully will last a little longer than the previous one! Anyway I am > very > glad that this turned out to be unrelated to the operating system, > although > I am still trying to get to the bottom of what happened to the other > machine > - which was clearly a software issue. > ---- hardly surprising at all...consumer grade SATA hard drives have become unreliable and that is why none of the manufacturer's warranties them beyond a year now.

For your first comment about the response of the system to a mistyped password in kdm I was not aware of - I'll ask if this it was a second login attempt when I talk to her later - and yes i do have kdm as the login manager. The status of the bz you referred to linked to an upstream bug which has not had any reply since the end of August! On the suggestions for trying to boot a livecd - I did try and after the POST, it gave "port 00:" at the top of the screen and a line of "stuff" - but did nothing further. I now have a new replacement HD which I will install over the weekend and see if I can then boot a liveCD to configure it - I'll report back in due course.