General :: Swap - Configure To Call Oom_kill On Its Own?
Jun 18, 2011
I do not want my computer to swap data to disk. I have no swap partition:
[Code]...
Back in the day (maybe as recently as kernel 2.4?) this used to work. Memory-hungry processes would be killed by the oom killer and I would restart them. But now (Linux 2.6.38-8-generic #42-Ubuntu SMP Mon Apr 11 03:31:24 UTC 2011 x86_64) google-chrome (13.0.782.24 beta) regularly sends my machine into a death-spiral of swapping. Or at least something that feels like swapping: X windows take forever to update, disk drive whirs, gnome panel memory chart hits the ceiling, and I see this message in the syslog:
[Code]...
But, according to syslog, the kernel does not run the oom killer. For good measure, I set swappiness to 0:
[Code]...
My workaround is to use ctrl-alt-sysrq-f when this starts to happen. Anyone have a recipe for configuring Linux to run oom_kill on its own, in this situation?
During the second OS (CentOS) setup, the shared SWAP partition was formatted and now I get the following under Debian: Code: # free total used free shared buffers cached Mem: 507524 251436 256088 0 11488 78332 -/+ buffers/cache: 161616 345908 Swap: 0 0 0 How to configure Squeeze to use again this SWAP partition?
In practice I have a script that call a java program that call a linux system command. The script if I run it, from a shell functions well,so it is not a java problem. The problem come out when i put this script in a crontab schedulation. The result in this case is that java do not execute the system command. I think it depends on crontab
Lucid on an Acer Travelmate800.Can anyone tell me why I have 0k for swap space? I allocated swap which I can see in my Disk Utility's 'volumes' display.
RAM for older machines like I use is fairly cheap these days. But flash memory is just as cheap or cheaper. So I'd like to ask about the feasibility of expanding my system's memory using flash memory. And about whether creating a partition for swap on the flash memory, or whether a swap file on the flash device, is the better way to go.
By flash memory I have in mind mainly USB sticks or what are sometimes called "pen drives." But I do also have CF and SD cards that, with the proper cheap adapter (one of which I already own for adapting CF) could be used to create extra swap space. So, what is the current consensus on the feasibility/advisability of using flash memory for swap? I've read about the limited write cycles of flash being an argument against using it for swap. But recent reading indicates to me that the limited write cycles problem applies mostly to older, smaller-capacity flash memory. Some will come out and say that, for larger-capacity flash memory, the life of the device is likely to exceed the amount of time your current computer will be useful (I think I've seen estimates in the range of 3-4 years life--minimum--for newer, higher-capacity flash memory).
A more persuasive argument I've heard against using flash memory for swap is that access times for these devices can be much slower than SATA, and maybe even IDE, hard drives. That would certainly dictate against using flash memory for swap.
So, how about some input on this issue? Anyone using flash memory for swap? If so, what kind (e.g., usb stick or SD/CF)? Are you using a swap file or a swap partition? How's system performance? Likewise, has anyone had flash-memory-used-as-swap die on them? The consequences would undoubtedly be dire. Also, has anyone measured flash memory access times to confirm or refute claims about slow access times? Are some types of flash memory better/worse than others in terms of access times?
i want to know how to call the other file in linux. example i got this code that store into example1.sh and i want to store the data into other file like example2.
what is what is signal 0 in linux system call ?i can't find that in kill -l signal list . i need to know what is signal 0 . * for examples signal 1 is SIGHUP and signal 2 is SIGINIT . what is signal 0 ?
I am writing a shell script lets say run.sh and i want to call another script lets say test.sh having permissions " root:sys" whereas the permissions for run.sh are "rootther".Problem is that i cannot change the permissions of either of them. run.sh is performing some other important functions as well and test.sh is a system script and i am not supposed to change that.If i do not change the permissions, run.sh simply skips the part where it is suppose to call test.sh.Can anyone tell me how do i call test.sh in run.sh without changing permissions of any of them.
I had an interview , where the interviewer asked me the question "How to create and call a process in unix". I answered that we can use the command FORK()/ CLONE() to create a process. But I didn't have the answer for the second question, how to call a process.
When I an in Emacs, I want to call info pages like in terminal. E.g., info date will call the date info page, but when I in Emacs, I can only use M-x info call the complete info pages, but I only want to see the date info page. Then which command I should execute in emacs.
I want to launch other executable from my cpp code. I am aware that system() call may do this , But, problem is that , I want my cpp code to capture the standard hat "called" program is writing to stdout.
I have 3 scripts that need to be run from the cd, I can run them separately but thought it would be easier to just call them from a single script.When I run the script it just goes straight to reboot, which is the last command. Can you see what I am doing wrong? I am logged in as root when I execute it.
I can find all kinds of information on calling a script from a makefile but not the other way around.I'm writing a VERY basic shell script, and I need to call a makefile located in a directory lower than the shell script but can't seem to figure it out../dir/make and ./dir/makefile don't work.
Here is a test code for memory alloc and free. It is to detect whether the allocs and frees match. From the result, after backtrace call, there is some memory alloc, but it does not free these memories after calling or the process exit. Why? Nomorlly, the memory alloced need to be freed after use, or the memory will leak. But I run this program for about some times, and find that though the system memory does not real leak,the free memory decreases, and the cache increases(Use free command).
I have Linpus runnig on my Aspire One. I would like to setup an icon on the menu to start a windows program using wine. I can start it manually and it works fine. I want to be able to access it directly from the desktop.
I have added a new system call for kernel version Linux 2.6.25-14.fc9.i686. after kernel compilation and running the user mode program i am getting the below error
Error :: Function not implemented Error number 29
I cannot find the syscall_table.S or entry.S file in the source code , so i didn't made the entry of the syscall in this file, how much necessary the entry of syscall presence in this list and how to get this file?
i using Popen() sys call in my code to run a script. i know it will return file pointer. but requirement is some how i need to get return status of process which is invoked by popen. there any possibility to get its status, if so please let me know. FYI i cant use system() here becoz of some limitations
I am searching for any system call similar to "ls" command we use in shell. My requirement is knowing the files and directories in curent working directory and process them based on there type. Here as of now I spawn another procees with system command like system ("sh ls -l | grep ^d | awk '{print $9}'").
Instead of this I want to use any system command where I can capture this information directly into my local character buffer. My opinion is that system calls will not spawn another process as a result less time it takes, another reason is once I use the system command again I need to capture the information to a local file then again read it into local buffer. I want to avoid the file manipulation here.
How can it be determined whether a function defined in some file of an application source-code(meant for multiple operating systems e.g., Linux, AmigaOS) is actually called when the application run on Linux?
As per my knowledge, one can use systemcall to call the driver functions for example using open, close... Now is there any way to call driver functions directly from application without using system call?
I have to run "pppd call idea" command from root shell every time to connect to internet from mobile. Now I want a script so that I just run it to connect. Something like :
#!/bin/sh echo "password" --stdin | su - pppd call idea
But its showing error that "standard in must be a tty". Why is this. Using CentOS 5.5
