I'm planning on setting up a new Linux box expressly for distributed computing (BOINC, SETI@home, etc.). All things being equal, what's better- More clock cycles or more cores?
View 4 Replieshow to show all program running (( computing activities )) on ubuntu 11.04 , like mac or this picture [URL]
View 5 Replies View Related"Servers aren't meant to have GUIs because they are a serious waste of CPU clock cycles." I encountered this line from somewhere here in ubuntu forums, but how could I install tools like mysql wrokbench and stuff, w/c will make my life a lot easier as an administrator? or is there such thing as remote administration?
View 4 Replies View RelatedLast year I was looking into fault-tolerant distributed file systems and I recall one kernel-based system that required a physical partition on each machine in the cluster, but would treat it as a single volume - ie. a write on one server would appear on the disk on all the servers.Unfortunately I didn't bookmark the specific system I was looking at, and now a year later I can't remember the details.What I don't want is NFS - a single file server with a file system mounted on various machines. What I do want is mirroring - one disk shared among multiple servers, so that if one server dies, it doesn't make any difference to the rest of them.
A bit of investigation turned up Red Hat's GFS, which kind of looks like what I want, but looks more and more like an NFS model to me. I was wondering what everyone's opinion of the various options out there were.
I have gotten my DNS servers up and running, and query against them using dig returns the proper information. How do I get the information to be distributed across the Internet now?
View 1 Replies View RelatedI am bit curious about Cluster Computing. I don't know much about it.Here is Some thing which I wanted to know.1) How can I configure fedora for Cluster computing .2) Is it possible to connect computer of different configuration(like AMD + Intel )? and also for different architecture ( like i386 + x86_64 ) for cluster computing.
View 7 Replies View RelatedWanting to evolve current cluster to GPU-enabled and retain ubuntu (champion the cause!) Details below; ill share all results/successes here on the forum for any interested. How best to get started ? Heterogeneous cluster of 40 quads/dual-quads w/4,8,16Gb memory All running ubuntu 8.04 or 9.04 Quantum Mechanical computations supporting neutron and muon beam experiments on red-wine/green-tea anti-oxidants, etc... Success of this centure would be precision independent (for now) single-precision floats are fine for now. I am more concerned with getting it going and making compatible with current codes - less concerned with accuracy of results for now. Double prcision floats will come anyways with the Fermis.
View 5 Replies View RelatedI have 2 nvidia gfx cards, if I use the nvidia settings tool to view 2 monitors with the same gfx card works fine, when i use one gfx card for one monitor and the other gfx card for the other monitor X doesnt start and just cycles through monitor flashes.
View 2 Replies View RelatedI was running 10.04 until yesterday, when it occured to me that I could upgrade to 10.10. So I went to Software Center, set it to get normal releases and left it to do its job. The upgrade appeared to go without a hitch and I rebooted. The login screen appeared. But just before I could click on my username and enter my password, the screen went blank and a second later the login screen was back. But then just before I could click... Undeterred, after half a minute of frantic clicking I did manage to click on my username and get the password prompt. This time, the login screen didn't go anywhere. Yay. To cut a long story short, this is now my standard logon procedure. However, the plot thickens. I appears that instead of 10.10, I ended up with 11.04, Natty Narwhal, which 'was released in April 2011'. If I download an .iso of Maverick and install it over my current version, will it leave my data unharmed AND reset everything so that it works again?
View 5 Replies View RelatedI'm trying to understand how deb packages work. Are there source debs?I assumed there were, but from[URL].. I see that source is distributed in a GZipped file.
View 3 Replies View RelatedI'm looking for the best distributed filesystem to share my home directory among all my computers, and I'm interested in what advice I could get here.
Here are my requirements:
1) It must be completely POSIX compliant, because I've noticed that many programs will break if not.
1a) Must support file locking (i.e. flock()) preferably sharing lock advisories between the clients
1b) Must support hard links, and be able to recognize a hard link connection by looking at the number of references
2) It must have encryption, and security suitable for it to be exposed to the internet.
2a) I would like to use password-less authentication. Preferably SSL client certs (HTTPS style). Static keys (SSH style) are also fine. I would begrudgingly accept a password authenticated system if it used a fairly secure challenge/response algorithm if it met all of my other requirements. (Extra points if it supports PKCS#11 and thus supports smart cards)
2b) Encrypted traffic must be an option.
2c) No thanks to kerberos. It was too annoying to maintain when I tried it before and I don't like passwords.
3) Preferably, it should be stable and fairly easy to maintain.
4) Optionally, it should be tolerant of network drops, and reconnect automatically as needed, because my laptop wireless is a little flaky sometimes.
I'm built an distributed application, and now I want to take some metrics to show how my application works. These metrics includes the time of computation, the time of communication and the amdhal law. I would like to know what tests can be done to show the performance of the application. So, I'm searching for a book or document that shows me what type of tests can be done to show the performance of an application.
View 2 Replies View RelatedA bunch of hosts all connected to each other.I'd like to have a directory where a group of hosts can dump files and all access, synchronising with each other their changes automatically and transparent to the user.How can I have it so that rsync updates both ways?
* host A asks host B for a list of files, modification dates, checksums
* then applies the rules to select which files it will fetch from B
--> if A owns the file then do not get modifications from B
--> files differ, then select newer file
* then fetches them
* and so on for each host...
I have a machine running Fedora 14 and a bunch of movies stored on it as '.iso' images. It is connected to my home theater. I used to use VLC to play these movies and it worked great for a long time. About 4 months ago, about an hour or hour-and-a-half into a movie, the audio suddenly "disappears" and the machine goes CRAZY with almost 100% of the CPU spent writing to the SWAP. At first I thought the machine was locked, but it isn't; it's so doggone busy writing swap. I am unable to get enough cycles to terminate anything. I found the swap activity through System Monitor - it was ABRUPT. Sound stopped, machine became preoccupied with swap.
I have removed/reinstalled VLC, the machine has undergone a couple kernel updates, and I have removed/reinstalled a number of things associated with audio (CD ripper, mpeg stuff, etc.) yet the problem persists. I don't know what happened or when (update-wise). Any body got any ideas? While a solution would be great, I'd also be happy with a couple decent suggestions on what to look for.
I was wondering, being a former Ubuntu user I was able to do cloud computing using Ubuntu One. Is there anything similar in OpenSUSE?
View 9 Replies View RelatedNever seen this happen before. I'm using the blank screen saver.
View 4 Replies View RelatedI've somehow got it into my head that it's possible to share CPU cycles, though I've no idea where from.So basically that's what I'm asking - is it actually possible to tell one system to 'donate' it's unused CPU time, cycles, whatever they are, for another's use?
View 4 Replies View Relatedwhat is High Performance computing and its uses.
what is active active and passive passive cluster ?
main()
{
unsigned long long tsc_start, tsc_end;
tsc_start = get_tsc();
// execute some code
[code]....
I just installed ubuntu 10.04.02 on to my tablet, aside from the dreaded intel video problem. Install went ok, but I'm finding now. That when I rotate the screen, the touch screen isn't adjusting to match. I'm not really ubuntu savvy.
View 6 Replies View RelatedI have a fedora 10 live CD currently i m running win XP with SP2 with the following configuration
Pentium 4 2.4 GHz
512 MB of Ram
160 GB HDD
INTEL d845 GVSR mother board
When I pop the live CD and reboot my system, the fedora 10 starts loading, it comes onto a select proper keyboard and after I clicked on the Next button tha app hangs. I used distributed CD and loaded iso images - result the same.
I'm using a program on a group of servers which contact each other using SSH. I've got the authentication keys setup with no password but when one box tries to connect to another it wants the password for the user it's trying to communicate with. So the main problem is there is a lot of communicating within this program so it's important to allow each box communicate with each other without the use of a password. I'm using OpenSSH 4.3 with OpenSSL 0.9.8 on Red Hat Enterprise Linux Server 5.3 (Tikanga). I'm also using SSH on Ubuntu(Intrepid) to connect to these servers at the mo as i don't have direct access to the boxes.
View 4 Replies View RelatedI was asked to research if it is possible to install a Debian distribution of Linux to HP Z420 Workstations as second OS (dual boot) in our lab. The lab is normally used for Computer Aided Design (Windows 7). The plan is to have parallel computing possible with these workstations on Linux side.
The link for HP Z420 Specs suggests amd64 distrubition should be supported. Tell me If I'm wrong as I have always been interested in software side of things.
Looking for info regarding installation of Debian to HP Z420 workstations as well as using these computers for parallel computing with Debian.
I have a couple of question regarding the screen blanker on Gnome desktop.I used to use a 1024x768 display with previous openSUSE distribution. With 11.3, I discovered the new "auto-configure" X feature. The default screen mode was 1600x1200, but I changed it to 1280x1024.My gfx board is a Matrox G400 DH. Hardware acceleration is disable because of a missing mga_dri.so (fall back to software rendering).I find some screen blanker modules are using almost all the CPU cycles. Animations are very slow, and it can take long before a keyboard hit or mouse movement makes to leave the blanker.
So the questions:- Is there a way to define another (smaller) screen resolution just for the blanker ?- Who should I try to convince to add mga_dri.so back in Mesa again ?- When the monitor go to sleep (DPMS), the blanker still is running, uselessly consuming CPU times. (I can see that because at the first mouse/kbd event, the monitor wake up and shows the blanker running.) Is there a way to configure the blanker to stop running when the monitor is sleeping ?- There are some modules which load images from HD (not the diaporama which load images in ~/images). But the shown image is always the default built-in one. Where is the blanker trying to load images from ?
I am running open suse 11.3 and keep up on maintenance. Ever since upgrading to 11.3 I find that the number of cpu cycles is being eaten for apparently nothing. In looking at the system monitor I frequently find that Xorg is using 24% and frequently more than that. What can be done to reduce the cpu cycles, or fix the problem?
View 9 Replies View RelatedI came across some problems. I need to conduct parallel computing on CUDA. I input "uname -a"and read these: "Linux BXSJC 2.6.38.8-35.fc15.i686 #1 SMP Wed Jul 6 14:46:26 UTC 2011 i686 i686 i386 GNU/Linux". I use CUDAtoolkit 4.0. when I input: cd /home/bxsjc NVIDIA_GPU_Computing_SDK/C and "make -i ",the computer returns
[Code]...
I have a problem in Ubuntu 10.04. The bug is well know which makes the hard disk head to park often (2-3 times per minute) that's dangerous for the drive in the long term and annoying for me (click-click-click).I found out the "ugly-fix" for old Ubuntu version which was :hdparm -B 254 /dev/sda instead of 128. It works. The problem is that it doesn't remains (when restart/standby/ac connection-disconnection).I found a script well known too :Code:1) make a file named "99-hdd-spin-fix.sh". The important thing is starting with "99".2) make sure the file contains the following 2 lines (fix it if you have PATA HDD):
#!/bin/sh
hdparm -B 255 /dev/sda
3) copy this file to 3 locations:
[code]....
I want to know the details about the implementation of distributed firewall in a local area network
View 5 Replies View Relatedis there a way to set-up a file system, where you have multiple nodes (PC's) linked to one storage pool? ZFS seems to support this idea with multiple disks on one node, but could you connect multiple nodes in this manner?
View 3 Replies View RelatedI am currently looking for a file system that needs to be distributed over several nodes and need redundancy, like RAID5, and the ability to grow if needed, like LVM. Also not all nodes are going to be located in the same data center, but I guess that's not that important as long as the connection between DCs is sufficiently fast. I am currently looking at AFS, Coda, GFS and OCFS to see if they have what I need.
Edit: I just figured that it may be better to have the RAID5-like setup within one location, and have RAID1-like mirroring between the locations. That would probably better for performance, right?