Operating Systems and Crashes: Why is the baby thrown out with the bath water?
Linux - GeneralThis Linux forum is for general Linux questions and discussion.
If it is Linux Related and doesn't seem to fit in any other forum then this is the place.
Notices
Welcome to LinuxQuestions.org, a friendly and active Linux Community.
You are currently viewing LQ as a guest. By joining our community you will have the ability to post topics, receive our newsletter, use the advanced search, subscribe to threads and access many other special features. Registration is quick, simple and absolutely free. Join our community today!
Note that registered members see fewer ads, and ContentLink is completely disabled once you log in.
If you have any problems with the registration process or your account login, please contact us. If you need to reset your password, click here.
Having a problem logging in? Please visit this page to clear all LQ-related cookies.
Get a virtual cloud desktop with the Linux distro that you want in less than five minutes with Shells! With over 10 pre-installed distros to choose from, the worry-free installation life is here! Whether you are a digital nomad or just looking for flexibility, Shells can put your Linux machine on the device that you want to use.
Exclusive for LQ members, get up to 45% off per month. Click here for more info.
Distribution: RPM Distros,Mostly Mandrake Forks;Drake Tools/Utilities all the way!GO MAGEIA!!!
Posts: 986
Rep:
Operating Systems and Crashes: Why is the baby thrown out with the bath water?
With all the infrastructure like Virtual Memory and the Virtual File System,why does the entire ram have to be dumped when there is a problem? Can't something monitor which application causes the crash, save the configruation and or environmenta variables and restart the process its self.
I don't know if a watch-dog timer could be implemented but it works on smaller systems.
Distribution: Debian, Ubuntu, CentOS; in days past Fedora, Solaris, SunOS, 4.2BSD, 4.3BSD, SVR4, AIX, HP-UX
Posts: 101
Rep:
If an application exits abnormally, the linux kernel rarely goes with it. Kernel panics and oopses occur, but rarely due to an application space problem. Do you have a specific instance of the problem?
Yeah, I was thinking the same thing ... there must be a specific issue, because most app errors end in a seg fault without crashing the system. So, do state the specifics ...
Distribution: RPM Distros,Mostly Mandrake Forks;Drake Tools/Utilities all the way!GO MAGEIA!!!
Posts: 986
Original Poster
Rep:
I would really rather know why crashes are not made obsolete.
List of my crashes
Application Crashes
Firefox: Windows don't stay loaded. If you move a window it blurs.If you hide the desktop windows and then reverse it, they will be transparent and you will see right through them to the desktop. Only the border will be visible.Part of a Browser window will act independently of the actual window that spawned it.If you move the window it stays where it was first created.
I assume this is because of javascript issues.
Movie Player crashes.It won't open a file. Some times I have to reboot for it to work again.
Linux Crashes: This is actually a dameon crash problably. The icons in my top tray of Gnome don't work and the Browser window icons in the bottom tray won't work either. You can't kill process using the GUI because you can't access the system monitor. This always seems to be related to another process but I am not positive. I should find out what the hot-keys are for the System Monitor.
Last edited by theKbStockpiler; 09-24-2010 at 02:55 PM.
Distribution: Debian, Ubuntu, CentOS; in days past Fedora, Solaris, SunOS, 4.2BSD, 4.3BSD, SVR4, AIX, HP-UX
Posts: 101
Rep:
Quote:
I would really rather know why crashes are not made obsolete.
This one I can answer with absolute certainty. Crashes are not made obsolete because humans are fallible, and their creations are fallible. Years ago, I had a book (maybe Structured Design by Larry Constantine) the glossary of which defined the word "debugged" as a hypothetical state of software. 30 years later, I've never seen a counter-example.
Hardware eventually breaks, Software eventually works
As for the OP, I don't see anything there that indicates a "Linux" (i.e. kernel) crash. X and associated candy ain't - although with KMS that might now be debatable.
Distribution: RPM Distros,Mostly Mandrake Forks;Drake Tools/Utilities all the way!GO MAGEIA!!!
Posts: 986
Original Poster
Rep:
Why can't a program un-crash by defult? Why are people f#$K ups
If there were such an "Application Restorer" it would not solve the applications problems but you would not have to reboot either. I am inclined to think there is a better solution than hitting the main power switch. I was interested to know if such a thing would only work if the O.S had to be turned inside out or something. Is the hurdle political or physical?
It's like building houses of non-flamable material OR having a fire hydrant every one thousand feet. You have to have one or the other. I understand that with the current accepted mindset that crashes will be a re occurring thing. I personally would like to write opensource apps and be on a board or something for Gnome or GTK so it' not like I'm non appreciative of my Linux-GLP privileges. I'm not a Draconian and understand that individuals are effected by the situations that they are in and situations that they can not control.
If you are familiar with Ff on windows you are aware that it has a dialog box that states that there has been a problem. So we hope that Ff actually has something that monitors issues and chooses to close the applicaiton in a controlled way. I have to believe that this is part of a M$ image thing so it's point and click users don't get confused. But along with the monitoring software it could save the info that is already in Ram and not start from scratch. I almost always suspend my computer because I keep a lot of webpages open. Not on the same monitor but I open closed ones in the tray when i need them. If I am reading about Bash I may have fifteen tabs on Bash on one Ff Window. So if Ff crashes it takes a while to reload sixtyfive pages or so. Appliction memory and data have their own addresses so it seems unnecessary to dump the Webpage images and such. There must be a reason why all data is dumped along with the application code.
Thank you for the reply!
Last edited by theKbStockpiler; 09-24-2010 at 10:12 PM.
Reason: I knew how to spell that!
if it wasn't a problem before and has only recently become a problem, how old is your motherboard? I had some leaky capacitors on a mother board and that cause a lot of strange issues. For some reason, Firefox really liked to instigate the issues. Maybe it was just me... but old hardware does fail in unexpected ways. Like a graphics card that starts to go bad my run everything fine until you try to play video. I had one that started playing video all corrupted. Additionally my x server would crash because of the graphic effects but windows would run fine because it was all disabled.
But if you are concerned about a rogue program crashing and taking up 100% of your cpu usage, there are ways to control how much each user on the system is aloud to take up. Try "man quota" and related manuals.
Anyway, there are lots of reasons for things to go wrong. If you find a bug and consistent steps to reproduce, then you report it so that it can, hopefully, get fixed.
You need to tell everybody you know if you have ideas on how to distinguish a crash/bug/resource intensive glitch in one program and a valid resource intensive task, say Gaussian blur on a really large image, in another program. I'm sure they would like to know.
Distribution: Debian, Ubuntu, CentOS; in days past Fedora, Solaris, SunOS, 4.2BSD, 4.3BSD, SVR4, AIX, HP-UX
Posts: 101
Rep:
Quote:
"If you are familiar with Ff on windows "
I, for one, am not. I was a Unix guy when DOS was a pup. I never was a competent windows user, never interested in being one. Although my current employer puts Vista box on my desk, and I do use it casually.
Quote:
like building houses of non-flamable material OR having a fire hydrant every one thousand feet
And yet houses still burn, and fire departments stay busy. I go back to my comments about human fallibility.
Quote:
If there were such an "Application Restorer" it would not solve the applications problems but you would not have to reboot either
I literally never have to reboot due to an application error. I'm not saying it can't happen that an application problem makes the kernel unusable, but it is outside of my experience.
I am sympathetic with your problems, but I suspect that your system is somehow misconfigured. Earlier H_TexMex_H asked about your video hardware. Are you pretty certain your video drivers are in good shape?
Distribution: RPM Distros,Mostly Mandrake Forks;Drake Tools/Utilities all the way!GO MAGEIA!!!
Posts: 986
Original Poster
Rep:
Group response
I don't have a graphics card at all.I just have the chip mounted directly on the board. It is an Abit (LG-95) board and about five years old. I have no idea about the video drivers unless it is shared with windows which is what I got with the board.(Is this assumption correct)? This sounds like something I should look into.
I never had the image problem on XP Pro with browsers but it does something simular with PDF files when you load them directly on the Browser and not to file. If I move a window over the loading PDF in windows the thing (usually another window) will leave a trail of images of itself on the window in the background.
If I download videos on Linux sometimes the image is all messed up at certain parts of the video with Totem. If I play it with VLC it is Okay. This stuff never happens with Nautilus windows.
Distribution: Debian, Ubuntu, CentOS; in days past Fedora, Solaris, SunOS, 4.2BSD, 4.3BSD, SVR4, AIX, HP-UX
Posts: 101
Rep:
You have refined your idea of the problem, and I think you should start a new thread to deal with your video questions. This thread, while interesting, is really not converging on a solution. The thread title and the General category aren't helping you. Start a thread in Linux - Hardware with a good title and a clearly stated question and I'd bet you'll have a solution before you know it.
Quote:
video drivers unless it is shared with windows which is what I got with the board.(Is this assumption correct)?
Linux video drivers are not shared with Windows.
Quote:
never had the image problem on XP
Continuing to work from a Windows context may not serve you well. First, most people here are either Linux enthusiasts or Linux professionals and may not know or care that much about Windows. That's my circumstance, anyway. Second, many Windows concepts don't apply very well in Linux.
I don't have a graphics card at all.I just have the chip mounted directly on the board. It is an Abit (LG-95) board and about five years old. I have no idea about the video drivers unless it is shared with windows which is what I got with the board.(Is this assumption correct)? This sounds like something I should look into.
Distribution: RPM Distros,Mostly Mandrake Forks;Drake Tools/Utilities all the way!GO MAGEIA!!!
Posts: 986
Original Poster
Rep:
Thanks for the push?
Quote:
Originally Posted by fuubar2003
Bingo.
I'll research bad graphics cards and see what comes up. I think I should also research "Crashes" and how Ff knows when to exit before a Windows user notices it is misbehaving. My own oppinion is if the application knows it is having problems why can't it salvange the data and just reload the code? Doesn't a Journaled File System do someting on theses lines? BTW I use the term FUBAR at work because it is more polite than stating that I work in a "Snowballing Cluster F#$K". Or the just a "Snowballing Cluster".
My own oppinion is if the application knows it is having problems why can't it salvange the data and just reload the code?
Once an application starts having problems it can't trust anything in its memory. An anology: a detective working on a case discovers that he's being dosed with hallucogenic substances. There's no way to be certain that the investigation hasn't been compromised, the best that can be done is pass along the notes to a sane person. Hopefully the detective hasn't completely destroyed the notes and actually handed them to the person he thought he did rather than some figment of his imagination.
LinuxQuestions.org is looking for people interested in writing
Editorials, Articles, Reviews, and more. If you'd like to contribute
content, let us know.