On Wed, 05 Jul 2006 07:58:29 -0500, Jeff Vian wrote:
On Wed, 2006-07-05 at 08:51 +0100, Paul Howarth wrote:
On Wed, 2006-07-05 at 12:11 +0800, Deepak Shrestha wrote:
Instead of updating driver from livna, this time I tried the
manual
installation of driver as instructed in nvidia forum http://www.nvnews.net/vbulletin/showthread.php?t=72490 its working fine now.
You should probably take a look at this URL:
That is old news.
There seems to be no date on that, but it seems I remember it from possibly as much as 6 months ago. I am not sure if the problems he described there continue to exist in new installs from nvidia since I have not installed that driver direct from nvidia on my FC5 system. I believe there have been 2 updated releases of the nvidia driver since his post of that.
He says that when using the driver from livna the problems noted are fixed so most won't have a problem. I personally never saw the problems he notes even when I used the driver directly from nvidia.
https://bugzilla.livna.org/show_bug.cgi?id=1069
status: RESOLVED resolution: WONTFIX
Why? proprietary. passing up the pipeline? too much work.
Thorsten's follow up commentary was additionally enlightening.
On Wednesday 05 July 2006 14:55, Scott R. Godin wrote:
He says that when using the driver from livna the problems noted are fixed so most won't have a problem. I personally never saw the problems he notes even when I used the driver directly from nvidia.
https://bugzilla.livna.org/show_bug.cgi?id=1069
status: RESOLVED resolution: WONTFIX
Why? proprietary. passing up the pipeline? too much work.
Thorsten's follow up commentary was additionally enlightening.
I'm curious. Just what is it, on a box that must be 100% stable, that needs the proprietary driver?
Anne
On Wed, 2006-07-05 at 15:56 +0100, Anne Wilson wrote:
On Wednesday 05 July 2006 14:55, Scott R. Godin wrote:
He says that when using the driver from livna the problems noted are fixed so most won't have a problem. I personally never saw the problems he notes even when I used the driver directly from nvidia.
https://bugzilla.livna.org/show_bug.cgi?id=1069
status: RESOLVED resolution: WONTFIX
Why? proprietary. passing up the pipeline? too much work.
Thorsten's follow up commentary was additionally enlightening.
I'm curious. Just what is it, on a box that must be 100% stable, that needs the proprietary driver?
Nothing. I can live fine without it, I just wanted 2d redraw accelerated, along with the openGL screensavers. I don't use the machine for games unless I've booted to windows first, (and that only rarely, as it's nothing to write home about cpu/vid wise), or for 3d-modeling.
It's got an nvidia chipset on the motherboard, so seeing that kmod_nvidia was available, built for my FC release, and my kernel, I went for it.
I didn't find out (or realize) until much much later just how unstable it really is.
Scott R. Godin schrieb:
On Wed, 05 Jul 2006 07:58:29 -0500, Jeff Vian wrote:
https://bugzilla.livna.org/show_bug.cgi?id=1069
status: RESOLVED resolution: WONTFIX
Why? proprietary. passing up the pipeline? too much work.
And didn't work.
Thorsten's follow up commentary was additionally enlightening.
Livna probably needs a group of maintainers for the driver. I want to get rid of maintaining the shit as soon as I can. I currently consider simply to stop maintaining and updating the driver -- maybe then some new people show up to maintain it.
CU thl
On Thu, 2006-07-06 at 11:48 +0200, Thorsten Leemhuis wrote:
Scott R. Godin schrieb:
On Wed, 05 Jul 2006 07:58:29 -0500, Jeff Vian wrote:
https://bugzilla.livna.org/show_bug.cgi?id=1069
status: RESOLVED resolution: WONTFIX
Why? proprietary. passing up the pipeline? too much work.
And didn't work.
Thorsten's follow up commentary was additionally enlightening.
Livna probably needs a group of maintainers for the driver. I want to get rid of maintaining the shit as soon as I can. I currently consider simply to stop maintaining and updating the driver -- maybe then some new people show up to maintain it.
Like I said, I feel your pain. It's got to be acutely so, when you take your own time, and apply it to something out of a desire to assist others, only to have your efforts thrown back at you or simply ignored when you attempt to pass feedback up the stream to the source of the problems.
It's frustrating for me, for the simple reason that I'd mostly just like the 2D basic screen refresh to be accelerated so that I could work somewhat more efficiently. (and having nifty-looking screensavers to entertain me occasionally) (and eventually perhaps having a few games to play of a more complex nature, like one of the Unreal series)
Faced with random gradual system corruption, on the other hand, I have little choice but to succumb to the inevitable.
The worst is having no conclusive, irrefutable proof in the form of kernel messages, that the nvidia driver is without a doubt the culprit of my recent experiences. Time will tell. I'd love to be proven wrong.
I miss 3dfx, and wish they hadn't lagged behind and been swallowed up by nvidia.
Thanks again for all your help. I realize this may not count for much, but I've been using the kmod_nvidia for quite a while now; on FC4 as well. Your efforts with easing the pain of installation paid off well -- This is one area that pleased me no end. :)
I guess the new updates of fedora 5 is ok. I have installed the nvidia driver ( I don't know about livna yet). so far, its behaving well, 2D and 3D are both working well.. What i did was
yum update install nvidia yum update again.
may be it could have worked if
yum update install driver from livna yum update again
I don't know why second update did some magic to my box.
I guess there are also other factors behind the corrpution of directories and other problems (may be bug in fedora itself). but so far i haven't seen any weiredness in my box yet as mentioned before like directory corruptions. Its working well and I am happy.
Thanks
On 7/6/06, Scott R. Godin scott.g@mhg2.com wrote:
On Thu, 2006-07-06 at 11:48 +0200, Thorsten Leemhuis wrote:
Scott R. Godin schrieb:
On Wed, 05 Jul 2006 07:58:29 -0500, Jeff Vian wrote:
https://bugzilla.livna.org/show_bug.cgi?id=1069
status: RESOLVED resolution: WONTFIX
Why? proprietary. passing up the pipeline? too much work.
And didn't work.
Thorsten's follow up commentary was additionally enlightening.
Livna probably needs a group of maintainers for the driver. I want to get rid of maintaining the shit as soon as I can. I currently consider simply to stop maintaining and updating the driver -- maybe then some new people show up to maintain it.
Like I said, I feel your pain. It's got to be acutely so, when you take your own time, and apply it to something out of a desire to assist others, only to have your efforts thrown back at you or simply ignored when you attempt to pass feedback up the stream to the source of the problems.
It's frustrating for me, for the simple reason that I'd mostly just like the 2D basic screen refresh to be accelerated so that I could work somewhat more efficiently. (and having nifty-looking screensavers to entertain me occasionally) (and eventually perhaps having a few games to play of a more complex nature, like one of the Unreal series)
Faced with random gradual system corruption, on the other hand, I have little choice but to succumb to the inevitable.
The worst is having no conclusive, irrefutable proof in the form of kernel messages, that the nvidia driver is without a doubt the culprit of my recent experiences. Time will tell. I'd love to be proven wrong.
I miss 3dfx, and wish they hadn't lagged behind and been swallowed up by nvidia.
Thanks again for all your help. I realize this may not count for much, but I've been using the kmod_nvidia for quite a while now; on FC4 as well. Your efforts with easing the pain of installation paid off well -- This is one area that pleased me no end. :)
-----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.4 (GNU/Linux)
iD8DBQBErPCOlfQeTCNIWLARAjsuAJ9zFi23rdVOqx4mxi0pvai5oUp+nACgk5MN WQsjCZIZKaL2xZFR94WjdCw= =lPis -----END PGP SIGNATURE-----
-- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
On Thu, 2006-07-06 at 22:36 +0800, Deepak Shrestha wrote:
I guess there are also other factors behind the corrpution of directories and other problems (may be bug in fedora itself). but so far i haven't seen any weiredness in my box yet as mentioned before like directory corruptions. Its working well and I am happy.
no, that's the thing. it doesn't show up right away. everything runs fine for weeks to months.. and then BOOM
What I'd HIGHLY recommend to you is this:
use tune2fs (manpage has details) to assign parameters to your various partitions so that they get checked at boot time automatically regularly. Why? well, the manpage points out:
"You should strongly consider the consequences of disabling mount-count-dependent checking entirely. Bad disk drives, cables, memory, and kernel bugs could all corrupt a filesystem without marking the filesystem dirty or in error. If you are using journaling on your filesystem, your filesystem will never be marked dirty, so it will not normally be checked. A filesystem error detected by the kernel will still force an fsck on the next reboot, but it may already be too late to prevent data loss at that point."
I, being paranoid, set mine thusly, for example (for my / partition among others):
tune2fs -c 25 /dev/sda3 tune2fs -i 3w /dev/sda3
so it gets automatically checked at boot time every 25 mounts or at most 3 weeks, whichever comes first. I did the same or similar with the rest of my partitions, excluding only the non-ext3 (windows, swap, etc.) partitions. If you never reboot, it'll never get checked, though. However if you experience a crash, this will help cover your ass by running fsck on the system if you're beyond the limit.
additionally, I set up a crontab entry to run rpm -Va once a week and report to me when things have changed.
(I've already set up /etc/aliases to forward root's mail to my user account, and ran newaliases to update the db)
(su -) to root (making sure to use '-' .. see man su for details why), and then crontab -e, and enter the following:
5 5 * * sun rpm -Va 2>&1
meaning: Run: rpm -Va 2>&1 At: 5:05am every Sunday
I'm still picking up the pieces.. today I discovered that /usr/bin/cpan had turned into a socket .. I still haven't replaced all the corrupted files with fresh installs of their respective rpms, but am slowly getting there a little bit at a time in between the work I'm doing.
good luck with it, but keep your eyes peeled for weirdness.
On 7/6/06, Scott R. Godin scott.g@mhg2.com wrote:
On Thu, 2006-07-06 at 22:36 +0800, Deepak Shrestha wrote:
I guess there are also other factors behind the corrpution of directories and other problems (may be bug in fedora itself). but so far i haven't seen any weiredness in my box yet as mentioned before like directory corruptions. Its working well and I am happy.
no, that's the thing. it doesn't show up right away. everything runs fine for weeks to months.. and then BOOM
What I'd HIGHLY recommend to you is this:
use tune2fs (manpage has details) to assign parameters to your various partitions so that they get checked at boot time automatically regularly. Why? well, the manpage points out:
"You should strongly consider the consequences of disabling mount-count-dependent checking entirely. Bad disk drives, cables, memory, and kernel bugs could all corrupt a filesystem without marking the filesystem dirty or in error. If you are using journaling on your filesystem, your filesystem will never be marked dirty, so it will not normally be checked. A filesystem error detected by the kernel will still force an fsck on the next reboot, but it may already be too late to prevent data loss at that point."
I, being paranoid, set mine thusly, for example (for my / partition among others):
tune2fs -c 25 /dev/sda3 tune2fs -i 3w /dev/sda3
so it gets automatically checked at boot time every 25 mounts or at most 3 weeks, whichever comes first. I did the same or similar with the rest of my partitions, excluding only the non-ext3 (windows, swap, etc.) partitions. If you never reboot, it'll never get checked, though. However if you experience a crash, this will help cover your ass by running fsck on the system if you're beyond the limit.
additionally, I set up a crontab entry to run rpm -Va once a week and report to me when things have changed.
(I've already set up /etc/aliases to forward root's mail to my user account, and ran newaliases to update the db)
(su -) to root (making sure to use '-' .. see man su for details why), and then crontab -e, and enter the following:
5 5 * * sun rpm -Va 2>&1
meaning: Run: rpm -Va 2>&1 At: 5:05am every Sunday
I'm still picking up the pieces.. today I discovered that /usr/bin/cpan had turned into a socket .. I still haven't replaced all the corrupted files with fresh installs of their respective rpms, but am slowly getting there a little bit at a time in between the work I'm doing.
good luck with it, but keep your eyes peeled for weirdness.
Don't you think its a tad bit odd, that no one else is reporting this mess? Look at the nvnews forums and see how many people are reporting filesystem corruption. You wont' find any, and there are alot of people reporting problems over there.
You stated that you ran memtest86, and it didn't find any problems. Did you check all the rest of your hardware? Maybe your CPU is bad? Maybe your disk(s) is failing. Maybe your system is overheating? Maybe there was a spike in your electricity?
Screaming fire in a crowded theater when your pants are on fire seems a bit ridiculous.
On Thu, 2006-07-06 at 08:05 -0700, Lonni J Friedman wrote:
Don't you think its a tad bit odd, that no one else is reporting this mess? Look at the nvnews forums and see how many people are reporting filesystem corruption. You wont' find any, and there are alot of people reporting problems over there.
You stated that you ran memtest86, and it didn't find any problems. Did you check all the rest of your hardware? Maybe your CPU is bad? Maybe your disk(s) is failing. Maybe your system is overheating? Maybe there was a spike in your electricity?
Screaming fire in a crowded theater when your pants are on fire seems a bit ridiculous.
The filesystem corruption I'm experiencing is a result of whatever wierdness was going on with writes from memory, and possibly swap (not 100% certain) but again, this is what happened *this time* (the fifth re-install of FC5) -- the four prior re-installs were caused by completely different problems. the sheer lack of any connecting events between five reinstalls is considerably annoying. kmod_nvidia is one of the connecting strings between all these events. I have now eliminated this thread and am waiting-and-seeing to determine if the problem still exists. so far, for a month, no sign.
the hard disk is brand spanking new, as is the SATA card. nice try though. and I, insane skewball that I am, did a low level format and deep badblocks check of the *brand new* drive, before doing ANYTHING with it. 250GB drive. took a looooooooong ass time. I'm pretty damned sure about the drive at this point.
overheating? nope. stuff wrong with the cpu? highly doubtful.
like I said, it's working fine now, and the only difference at all is the absence of kmod_nvidia.
filesystem corruption was just one of the many symptoms I experienced.
However, the advice I gave was sound. automatic checking of the filesystems at boot time is off by default, and as you can see, the tune2fs manpage does warn that this is probably not a good idea.
additionally running rpm -Va once a week in the wee hours isn't a bad idea either, if you're bullish on the integrity of the installed packages.
you could try toning your insulting tone down a bit. I've been very patient with you about it but you're wearing on my nerves gradually... If you're trying to get a rise out of me, save it.
*IF* the problems re-occur (and they might, I didn't re-install the system this time, just recovered via fsck, removed kmod_nvidia, and reinstalled a few of the badly corrupted essential rpm packages, although I haven't gotten to everything yet), I'm perfectly willing to concede that kmod_nvidia is not at fault. This has not yet happened. If it does, I'll happily investigate further for a real culprit.
So far, the finger still points quite firmly at kmod_nvidia, and it's becoming increasingly difficult to support any other avenues of culpability.
It's pure Holmsian deduction: "How often have I said to you that when you have eliminated the impossible, whatever remains, however improbable, must be the truth?" http://yoak.com/sherlock/stories/sign_four/sign_of_four.txt
Instead of fingerpointing and arm waving, how about you do something constructive, and instead of telling me how many people appear to be running just fine, either _PROVE ME WRONG_ with cold facts, or take *your* "personal FUD campain" somewhere else. Hm?
On Thu, 2006-07-06 at 11:53 -0400, Scott R. Godin wrote:
On Thu, 2006-07-06 at 08:05 -0700, Lonni J Friedman wrote:
Don't you think its a tad bit odd, that no one else is reporting this mess? Look at the nvnews forums and see how many people are reporting filesystem corruption. You wont' find any, and there are alot of people reporting problems over there.
You stated that you ran memtest86, and it didn't find any problems. Did you check all the rest of your hardware? Maybe your CPU is bad? Maybe your disk(s) is failing. Maybe your system is overheating? Maybe there was a spike in your electricity?
Screaming fire in a crowded theater when your pants are on fire seems a bit ridiculous.
The filesystem corruption I'm experiencing is a result of whatever wierdness was going on with writes from memory, and possibly swap (not 100% certain) but again, this is what happened *this time* (the fifth re-install of FC5) -- the four prior re-installs were caused by completely different problems. the sheer lack of any connecting events between five reinstalls is considerably annoying. kmod_nvidia is one of the connecting strings between all these events. I have now eliminated this thread and am waiting-and-seeing to determine if the problem still exists. so far, for a month, no sign.
the hard disk is brand spanking new, as is the SATA card. nice try though. and I, insane skewball that I am, did a low level format and deep badblocks check of the *brand new* drive, before doing ANYTHING with it. 250GB drive. took a looooooooong ass time. I'm pretty damned sure about the drive at this point.
overheating? nope. stuff wrong with the cpu? highly doubtful.
like I said, it's working fine now, and the only difference at all is the absence of kmod_nvidia.
filesystem corruption was just one of the many symptoms I experienced.
However, the advice I gave was sound. automatic checking of the filesystems at boot time is off by default, and as you can see, the tune2fs manpage does warn that this is probably not a good idea.
additionally running rpm -Va once a week in the wee hours isn't a bad idea either, if you're bullish on the integrity of the installed packages.
you could try toning your insulting tone down a bit. I've been very patient with you about it but you're wearing on my nerves gradually... If you're trying to get a rise out of me, save it.
*IF* the problems re-occur (and they might, I didn't re-install the system this time, just recovered via fsck, removed kmod_nvidia, and reinstalled a few of the badly corrupted essential rpm packages, although I haven't gotten to everything yet), I'm perfectly willing to concede that kmod_nvidia is not at fault. This has not yet happened. If it does, I'll happily investigate further for a real culprit.
So far, the finger still points quite firmly at kmod_nvidia, and it's becoming increasingly difficult to support any other avenues of culpability.
It's pure Holmsian deduction: "How often have I said to you that when you have eliminated the impossible, whatever remains, however improbable, must be the truth?" http://yoak.com/sherlock/stories/sign_four/sign_of_four.txt
Instead of fingerpointing and arm waving, how about you do something constructive, and instead of telling me how many people appear to be running just fine, either _PROVE ME WRONG_ with cold facts, or take *your* "personal FUD campain" somewhere else. Hm?
While I cannot prove beyond doubt that the kmod-nvidia package was not responsible for your problem, Lonni's point (and I concur) is that if the nvidia driver were the party at fault then others surely would be seeing the same problems you have. I have seen no one else reporting these problems and apparently Lonni has not either, in spite of my participation on several related mailing lists.
One thing that disturbs me is your insistence that your problems are related to a video driver when the problems you identify are all of a filesystem corruption type. I went through a period of about 3 months where I was getting filesystem corruption on a brand new drive. After I decided that it may have been related to problems with the drive and I contacted the manufacturer, it was determined the new drive was actually bad. (The OS never logged a single error on the drive, nor did smartd detect it.) I wound up replacing the same drive 4 times via RMA since it seemed they had a bad batch of drives and I kept getting the bad ones. Eventually the symptoms were eliminated and I now have a stable system. I had similar experiences when a power supply was marginal and caused flaky writes to the drive in that system.
You may need to consider other possibilities even though you insist the video driver is at fault.
My point is that your viewpoint seems a bit myopic and that since you have "identified" the problem you seem unwilling to consider any other possibilities. I wish you the best, but please take your vitriolic comments someplace else. If no one else is seeing the same problems in spite of using the same software then it seems very likely that you have picked an innocent victim and that your problems may surface in a different manner or resurface even though the "cause" has been removed.
I, personally, saw nothing wrong with Lonni's comment that triggered this tirade from you. I have seen you respond forcibly and with great heat to several who indicated you may be picking on the wrong cause.
https://www.redhat.com/archives/fedora-list/2006-July/msg00078.html says you have a sata problem (maybe responsible for filesystem corruption?) https://www.redhat.com/archives/fedora-list/2006-July/msg00486.html vitriol with no constructive value and more in the same tone. I did not look at any from June for this.
Please, please, just step back, take a deep breath, and think, instead of blowing up at the least little thing that is contrary to your beliefs. Vitriol does not win friends or support. Honesty, openness, and attempting to help yourself instead of blaming others does.
BTW, Something seriously is a problem with your timeline. You have said you have done 5 complete reinstalls with weeks or months between [1], and all were FC5 and all triggered by the nvidia driver. How did you get that many installs with that time period for each, between the release of FC5 on February 28 [2] and today, July 6? FC5 has only been out for 4 months total, so most would have HAD to be of less than 2 weeks each if you had even one that was over a month. Unless you are exaggerating???
[1] https://www.redhat.com/archives/fedora-list/2006-July/msg00627.html [2] http://fedora.redhat.com/docs/release-notes/fc5/
-- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
I was using yum to update the kernel, and although the first upgrade kept the original one, and gave me revision 2139, the one I did yesterday gave me 2145 but wiped out my original 2054..... I didn't see an option to keep old ones, what is the official way to keep the old ones?
- peter
On 7/7/06, Peter Teuben teuben@astro.umd.edu wrote:
I was using yum to update the kernel, and although the first upgrade kept the original one, and gave me revision 2139, the one I did yesterday gave me 2145 but wiped out my original 2054..... I didn't see an option to keep old ones, what is the official way to keep the old ones?
- peter
-- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
you have to edit the installonlyn.conf in /etc/yum/pluginconf.d directory.
example file [main] enabled=1 # this sets the number of package versions which are kept tokeep=5
change enable= to 0 to keep all kernels tokeep= is the number of prior kernels you would like to keep. I changed my system to keep 5 prior kernels.