Ubuntu Servers :: How To Torque On Ubuntu 10.04 On A Single Multicore Machine
Jun 17, 2010
I updated this on jul 21 using comments below Torque is a batch job queuing system that is used on clusters. But I find it handy to use it on my multi-core workstation as well. It allows jobs that need to be run to be schedule by multiple users. The scheduler will make sure that not too many jobs are run simultaneously which could cause high system loads or memory issues. I previously posted how to install torque on ubuntu hardy from the torque source package. However, torque is now in the repositories of lucid and here are the steps that I had to take to get it to work on my workstation.
For this setup I kept the server host name 'torqueserver' which is the default in the package. You can do the same or use a fully qualified domain name. In that case, you will have to adept the steps somewhat. My workstation has 8 cores, and I only want to give 6 of them to the que. Please adapt your numbers accordingly.
0) open root terminal
Code:
sudo -s
1) add torqueserver as an alias to /etc/hosts.
Code:
gedit /etc/hosts
change 127.0.1.1myHostName to 127.0.1.1myHostName torqueserver*) see post by drlemon.
Alternatively use a resolvable host name (check with 'host $HOSTNAME') in the file:
/var/lib/torque/server_name and whereever torqueserver is used below, use that host name
2) install torque from repositories
Code:
apt-get install torque*
[code]....
This works for me but probably requires more configuration in a demanding computing environment. Check out the torque website for more queue configurations, user management etc.
View 3 Replies
ADVERTISEMENT
Mar 17, 2011
I am experiencing a multicpre problem with fedora14 / kernel 2.6.35.6-45.fc14.x86_64 #1 SMP. (standard from the distro). The host has a total of 2 cores .yet a virtual machine started with 'smp 2' ( or xml : <vcpu>2</vcpu>) will not run on 2 pyhs cores . it will use just 1 core ('top' at the host shows 100%cpu while running an endless loop process on the VM) and libvirt 'virsh vcpuinfo <vm name>' shows only 1 physical cpu used (although 2 vcpus are used : 0, 1) The kernel seems to be an SMP one ( so says the uname -a . what does '#1 SMP' mean ?) . Am I missing something ? same XML on a DEBIAN machine will run the VM with 2 phys cores as should although the host there has 4 cores (>2)Should I recompile the kernel src with some addtional flag ?
View 7 Replies
View Related
Jun 1, 2010
I have this intra net server project going on and now I moved to 10.04 however there are still some things that I would like to see clarification and instructions on. I am interested to set up multiple parallel websites for my apache server, however I am not sure how to do this exactly. Now I have solid address rivera.wippies.net and port 80 redirecting to my server. What I would like to get done is that I get multiple independent of each other websites for my server I was thinking of making websites like this
/var/www/site1 (which would be as rivera.wippies.net)
/var/www/site2 (which would be as rivera.wippies.net/othersite)
/var/www/site3 (which would be rivera.wippies.net/secondothersite)
etc, so that I have multiple "individual" websties for my server. Requirements would be that each of these websites could have SSL encryption as needed available too, since some of the website could have confidential information.
View 9 Replies
View Related
Sep 29, 2010
I want to know the configuration steps for 2 mail servers on diferent IPs. But I want this setup on single vmware machine(debian).
View 3 Replies
View Related
Jan 3, 2010
I am running Ubuntu 9.10 x64 on a relatively old laptop and I find Grub is taking about 20 seconds to initially load before booting the system. The laptop only has a single 120GB hard drive, and I chose the "use entire disk" option when installing, so I do not have any fancy partitioning scheme.When the system boots I get the "LOADING GRUB" text displayed and then shortly after it just shows the flashing cursor with no text, and stays there for about 20 seconds. After that it finally starts loading the system.All the searching I have done only turned up solutions relevant to people with multiple hard drives. This was not an issue on Grub 1.
View 1 Replies
View Related
Feb 3, 2011
Can I have 3 drives in one physical server and treat each drive like a server?
If so, would I install the operating system on each drive, and give each drive an IP address?
I was wondering if I could run a webserver emailserver and dbserver on a single machine? Then later on, move the drives to their own hardware.
View 9 Replies
View Related
Jun 14, 2010
I always keep my headphones as well as my 2.1 speakers connected simultaneously on my PC (Gigabyte motherboard). Both these audio devices work fine simultaneously on my Windows 7, but in Ubuntu, only my headphones work, speakers don't work at all. I tried changing some settings in SOUND PREFERENCES as well as ALSA SOUND MIXER, but not luck so far.
View 1 Replies
View Related
Apr 16, 2010
I just installed Fedora 13 on my ESX box.I have Fedora 13 Machine which was early having 1 network adapter.I added a new Interface type: e1000 to this VM.Now,My ifconfig says:
Code:
[root@fedora-13 ~]# ifconfig
eth0 Link encap:Ethernet HWaddr 00:50:56:BA:00:15
inet6 addr: fe80::250:56ff:feba:15/64 Scope:Link
All i was trying to provide IP to eth0 and eth1 through dhcp.is it possible to provide two IP address to eth0 and eth1 both through DHCP.
View 4 Replies
View Related
Feb 2, 2010
Does anyone know if there is any way to configure 50K "virtual" IPv6 addresses on loopback device in Linux?
The aim is not to add all 50K IPv6 addresses one by one on the loopback/ETH device which will probably mess up the ip table on the system. In IPv4, I am able to achieve that by specifying the IP address subnet on loopback device (e.g: "ip addr add 10.1.0.0/16 dev lo"). The same command does not seem to work the same way for IPv6. It only adds a single IPv6 address on loopback device and it automatically adds an "unreachable route" entry on the ipv6 route table for that IPv6 network prefix.
The reason I need this is because I am working on an application which tries to simulate 50K IPv6 addresses on a single Linux box. The kernel version I am currently using is RHEL 2.6.9.55.
View 6 Replies
View Related
Jan 27, 2011
i wonder how i can update one single thing on my debian machine? without apt-get upgrade. lets say i just want to update my awstats install but not my php install, or my kernel version but not someting else, how can i do this ?
View 12 Replies
View Related
Dec 13, 2010
I am find myself in need of building what amounts to 3 ip networks on 1 physical lan.
first LAN is another companies network on the 10.0.0.0 network a mix of pc's and equipment
second is LAN new company network 192.168.168.0/24
third network is an ip phone system for both i would like to use 192.168.200.0/24
i am hoping to use the same dhcp server to do this, assigning address to the old company's 10. network via mac address reservation.
View 9 Replies
View Related
Oct 26, 2009
I have apache server running on Fedora 11. I want to test Self-signed certificate for SSL setup with Apache.HOw can I setup?
View 1 Replies
View Related
Apr 4, 2011
I have a network of 100 machines, all with ubuntu Linux. Is there a limit to the number of machines that can connect to one single machine (at the same time)? For example, can I have 99 of my machines maintain continuous ssh connection to the 100th machine? Can I have every one of my machines (every one of the 100) maintain a continuous ssh connection to all other 99 machines? How much memory does each such a connection take?
View 2 Replies
View Related
Mar 8, 2011
I have few virtual machines running on one of xen servers & I'm experiencing a problem booting up one of clients "domain".
How can I force this machine to boot up to runlevel 1? I can't see grub menu when first booting it up!
View 6 Replies
View Related
Jun 21, 2010
I have recently loaded the new LTS SERVER 10.04 release and it loaded without errors. My only issue is that it does not seem to correctly identify the Dual Cores for each processor. (The HP xw9300 I have has two hardware CPUs - each one being Dual-Core - so there should be a total of 4 CPUs recognized - unless I am mistaken?) How I can debug the processor recognition for multi-core support to get the additional cores recognized.
My /proc/cpuinfo file looks like this:
Code:
xxxx@hostname:~$ cat /proc/cpuinfo
processor : 0
vendor_id : AuthenticAMD
cpu family : 15
model : 37
model name : AMD Opteron(tm) Processor 248
stepping : 1
cpu MHz : 2194.452
cache size : 1024 KB .....
address sizes : 40 bits physical, 48 bits virtual
power management: ts fid vid ttp
View 1 Replies
View Related
Jun 23, 2010
If I have only 1 physical WLAN interface, is there some hackery that can be done so that it can be a client of an existing access point and at the same time also act as an access point for other clients? I have an existing 802.11g ADSL router, and I'm going to be building an HTPC which will have 802.11n. My laptop also has 802.11n, but at the moment it only connects at 54Mbps because thats what the AP supports. I'd like to be able to have the HTPC be a client of my ADSL router, but have my laptop be a client of the HTPC, so copying files to it will be faster.
View 1 Replies
View Related
Feb 16, 2010
What i meant was, in a multicore system with 'n' number of cpu's, will each cpu have its own linux kernel installed, or will all cpu's have a copy of the same kernel image? Another question regarding multicore is kernel tracing. Since tracing works at the kernel level, then should i perform tracing for each cpu's seperately or kernel tracing can be performed as a whole for all cpu's put-together?
View 3 Replies
View Related
Apr 8, 2010
I wish to know that which gcc release version had support on ARM 9 multicore system by using pthread library for its parallelizing method on ecos rtos.I assume that the compiler that support the single ARM 9 processor will surely compile its multicore version also.But i need a clarification about it.It is good if you give a article or provide a suitable link on the ARM multicore programming and its compilation method
View 1 Replies
View Related
May 3, 2011
I've got an intel i7 920 CPU. It has 4 cores, but thanks to hyper threading technology each core could run 2 threads so in /proc/cpuinfo it shows "8 processors".
1. what would be the optimal -j flag when compiling programs on my CPU. I've come across -j7. Why 7?
2. Where and what would I put this flag to make it permanent (eg. .bashrc?)?
View 7 Replies
View Related
Apr 2, 2010
I have a folder just for download files (e.g., application update files). I would like to send the Content-MD5 http header for each file a user downloads.So, is there a way to send ContentDigest header for the contents of a specific directory only?The apache documentation makes it sound like I can set a context so that it applies to a directory only, but I can't figure out how this is done if this really is an option.Can anybody tell me if I am interpreting the documentation correctly and, if so, how to set apache to work this way?
View 4 Replies
View Related
Apr 18, 2010
I recently set up a computer running Ubuntu 9.10 Desktop for use as a server, and successfully have it hosting game servers.I'm relatively new to Ubuntu, so I'm not sure how to go about this, but ideally I would like a user to be able to access a single folder on the server over the internet, and be able to upload and download files from it.It would be nice to set up multiple folders like this, so different people would have access to their own personal folder, but not anyone else's.
View 9 Replies
View Related
Jun 17, 2010
My server keeps displaying a message: "The computer needs to restart to finish installing updates. Please save your work before continuing" every time I log on. Last night I did four consecutive reboots but the message is still there. I log i remote with NX / Gnome. Synpatic shows "0 to install/update 0 to remove".Is this normal? Does every single update need an independent reboot, so there is a queue to catch up with? Or is there some reboot flag that is not resetting properly after reboot?
View 9 Replies
View Related
Jul 26, 2011
I'm using my home server to netboot into live CDs. To summarize that article, netbooting a liveCD system involves passing an NFS host:/path in kernel's command line, which is subsequently mounted as /cdrom on the live system. The article mentions copying the contents of the iso image to a place on the hard drive, but I see no reason for this, since the contents need to be read-only anyway, which is easy to achieve by mounting the iso through a loop device:
Code:
# mount
...
/dev/loop0 on /mnt/kubuntu-lucid-amd64 type iso9660 (ro)
/dev/loop1 on /mnt/kubuntu-natty-amd64 type iso9660 (ro)
[code]....
Now, I am using NFS to mount a regular (rw, HDD) directory which I use as a network drive and I can mount any subdirectory and still access all the files within, ie. when I export /data on the server, I can still mount /data/downloads on the client:
Code:
server:# exportfs
/data 192.168.1.4/32
client:$ sudo mount -t nfs 192.168.1.1:/data/downloads /mnt/netshare
[code]....
Is it possible to export just /mnt and be able to mount any/all of the subdirectories? NFS doesn't like going across multiple file systems:
Code:
server:# ls -al /mnt/kubuntu-natty-i386/
total 3532
dr-xr-xr-x 11 root root 4096 2011-04-27 11:29 .
drwxr-xr-x 7 root root 4096 2011-07-25 21:46 ..
[code].....
When I try to export just /mnt and mount /mnt/kubuntu-natty-i386 on the client, the mount succeeds but the mounted directory is empty.
View 1 Replies
View Related
Jul 21, 2010
I'm wondering if anyone is having a similar issue to what I'm experiencing. openSuSE 11.2 was very fast on my system. Since installing openSuSE 11.3, I've been experiencing many short term hangs where I'll lose control of the mouse for a few seconds (usually about 2-5) and the system will do nothing. I don't have unusual load during these time periods and I don't really see any spikes during these periods. If I leave a cpu monitor active, I can see a very very light spike one core0 when it happens, but the mouse and keyboard (and seemingly the whole desktop) hang during this spike so I can't see what's causing it, if that's really the culprit.
I am almost suspecting that I'm experiencing an issue with interrupts... since it seems that everything is hanging for a few seconds, then coming back to life. It does seem to happen more often when I'm resizing or minimizing windows... but that's about all I can say as it also seems pretty random.Here's some basic info on my hardware:ASUS M3A78 Pro motherboardAMD Phenom(tm) 9950 Quad-Core ProcessorGeForce GTX 260G15 Keyboard and MX revolution mouse (just in case USB interrupts turn out to be the issue)So, is anyone else experiencing a similar issue? Maybe we can narrow down what we have in common. Oh, I'm running x86_64.
View 9 Replies
View Related
Dec 8, 2010
I have been searching for the multicore scheduler support for ARM, in linux kernel.
CONFIG_SCHED_MC, CONFIG_SMP are there for x86..what about ARM ? will the same configs do the job...
View 1 Replies
View Related
May 20, 2010
I'm running Ubuntu Server 10.04 and have a secure (SSL/TLS) FTP server on it. However, I'd like to use this FTP server to update programs I made using Microsoft Visual Studio. Unfortunately, in Microsoft's infinite wisdom, secure FTP servers cannot be used. Rather than use an insecure FTP server, I want to set up my secure FTP server to be able to access whatever I need to on the machine, and then add an insecure FTP server that only has access to the directory where I put my update files. I am currently using vsftpd as my FTP server. Is there any way that I can set up two FTP servers on this single machine?
View 6 Replies
View Related
May 2, 2010
I want to fix my disks with fsck but using sudo init 1 is not allowing me in. It shuts my services down, then gives a "Give root password for maintenance" prompt. I've tried my password, but it rejects it.
View 3 Replies
View Related
May 5, 2010
I have an Ubuntu Server on (8.10) running under Citrix XenServer (though that shouldn't make a difference).
I installed on a single disk:
xvda1 - 200 MB - /boot
xvda2 - 9.8 GB - LVM (ubuntu-base)
The LVM is:
swap - 1.0 GB
root - 8.8 GB - /
I have successfully gotten this converted to RAID1 by adding a new drive (xvdb) and following the Debian howtoforge article [URL]
What I have not been able to do, is get grub working properly.
If I fail xvdb and reboot the system, everything comes up and I can reboot and run.
If I fail xvda and reboot the system, XenServer gives me a bootloader error. i.e.: no grub
If someone has done this, can they tell me what grub commands to run to get a successful boot of the primary disk fails?
View 1 Replies
View Related
Nov 16, 2010
I have a nice little interface that generates an exim config file and a script on the server to check for changes. Unfort I cant find where to set the config file location.
E.G. I have my script generate the file to /etc/exim/exim4.conf (could be /usr/share/foo/bar.conf just as easily)
How can I point exim in the direction of that file?
View 1 Replies
View Related
May 3, 2011
im on 10.10(desktop) and mdadm was v2.8.1 from 2008, very out of date so i tried 3.2.1 -> no change. mdadm raid1 read speeds are the same as single disk. note i used the tests in the disk utility benchmarking tool at first --these showed raid 5 atleast to be much better but when i tried dd reads raid5 dropped off with larger data to almost the same (slow) speed as raid1. compare:
[code]....
using two partitions will be enough to show raid1 performs at single disk speed. I dont really want to use a 4 disk raid0 just to get the read speed i should be able to get with raid1 as i dont really care about the size loss. I would of course use raid10 but i have found this suffers from the same problem (achieve same read speed as 2 disk stripe). So whilst im shocked others aren't reporting this, unless there is some obscure reason why my system would give these results i think raid1 in not behaving as it should.
View 6 Replies
View Related