Re: [one-users] Problem with OpenNebula debian package based setup.
Hi, I think your problem is when you ssh into localhost (frontend). Try this: ssh onea...@frontend It will probably complain about the host key. Fix it by removing the offending keys from the ~/.ssh/known_hosts file. Regards, Jaime On Tue, Dec 7, 2010 at 3:11 PM, KING LABS kinglabs...@gmail.com wrote: Yup I am able to login . On Tue, Dec 7, 2010 at 5:56 PM, Damien Raude-Morvan draz...@drazzib.comwrote: On Tue, 7 Dec 2010 16:00:03 +0530, KING LABS kinglabs...@gmail.com wrote: Hi All, Hi, So far I have been used express script to configure the setup, today i switched to debian to install 2.0.1 with the new deb package released yesterday. Configured exactly as given in the README file http://git.debian.org/?p=pkg-opennebula/opennebula.git;a=blob_plain;f=debian/opennebula.README.Debian;hb=HEAD From oneadmin account on frontend could try to run ssh onead...@192.168.138.231 ? Cheers, -- Damien ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] Open Nebula and Scheduling
Hi, a) Open Nebula can interface with a third-party scheduler to schedule the standing-up of VM instances or Yes, it's actually possible to swap out the default scheduler and replace it with a different scheduler. Have you taken a look at Haizea? (http://haizea.cs.uchicago.edu/) It is a more advanced scheduler (supporting advance reservations, queueing, etc.) that can be used as a scheduling backend with OpenNebula. Although, from what you've described, it might not meet some of you're needs, it might help you get a sense for what an alternate scheduler looks like. b) It would be possible to extend open nebula to interface with a third-party scheduler and which bits of the code base I should be interested in for this. With Haizea, I actually didn't have to make any changes to OpenNebula itself. Haizea runs as a separate daemon, and basically (1) periodically probes OpenNebula to ask if there are any new requests to take care of and (2) calls OpenNebula to start/stop/suspend/resume/etc. VMs. Some functionality that we're considering might require writing hooks for OpenNebula (http://opennebula.org/documentation:rel2.0:oned_conf#hook_system), but that still doesn't require modifying any of OpenNebula's core code. In Haizea, users still request the VMs to OpenNebula (and Haizea decides when/where to schedule them). From what you're describing, it sounds like you would like users to request jobs to an HPC scheduler, and then have the HPC scheduler provision resources using OpenNebula, right? Although this is different from the Haizea use case, my gut feeling is that this should still be possible as long as the scheduler provides pre- and post-job hooks so you can create the VMs and do any necessary cleanup. Cheers! -- Borja Sotomayor Scientific Writer, Computation Institute Lecturer, Department of Computer Science University of Chicago http://people.cs.uchicago.edu/~borja/ Community Manager, OpenNebula project http://www.opennebula.org/ ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
[one-users] onevm create failed
hi, now i have changed the suoders-file, but the error is still the same (onevm create failed). thanks for reply. note (on the execution-host): pwd: /srv/cloud/one which xm: (empty result) sudo which xm: (/usr/sbin/xm) xm list: (not enough privilegs) sudo xm list: (dom0 is seen) error-message: Wed Dec 8 17:12:18 2010 [VMM][I]: Command execution fail: /srv/cloud/one/lib/remotes/vmm/xen/deploy node1 /srv/cloud/one/var/30/deployment.0 Wed Dec 8 17:12:18 2010 [VMM][I]: STDERR follows. Wed Dec 8 17:12:18 2010 [VMM][I]: sudo: /usr/sbin/xm: command not found Wed Dec 8 17:12:18 2010 [VMM][I]: ExitCode: 1 Wed Dec 8 17:12:18 2010 [VMM][E]: Error deploying virtual machine Wed Dec 8 17:12:18 2010 [DiM][I]: New VM state is FAILED Infos: two hosts. debian lenny. xen hypervisor. one network-interface per host. master: 192.168.2.108 node: 192.168.2.102 ok: passwordless ssh-connection login as oneadmin user (already works) ok? # User privilege specification (on every machine the same) root ALL=(ALL) ALL oneadmin ALL=(ALL) NOPASSWD: ALL %xen, %cloud ALL = NOPASSWD: /usr/sbin/xm %xen, %cloud ALL = NOPASSWD: /usr/sbin/xentop ok? # network settings (frontend): auto lo iface lo inet loopback # The primary network interface allow-hotplug eth0 auto eth0 iface eth0 inet static address 192.168.2.108 netmask 255.255.255.0 network 192.168.2.0 broadcast 192.168.2.255 gateway 192.168.2.1 dns-nameservers 192.168.2.1 ok? # network settings (execution-host (node)): auto lo iface lo inet loopback allow-hotplug eth0 # The primary network interface auto eth0 iface eth0 inet static address 192.168.2.102 netmask 255.255.255.0 network 192.168.2.0 broadcast 192.168.2.255 gateway 192.168.2.108 dns-nameservers 192.168.2.108 # Primary (internal) cloud bridge interface auto br0 iface br0 inet static address 192.168.2.102 netmask 255.255.255.0 network 192.168.2.0 broadcast 192.168.2.255 gateway 192.168.2.108 dns-nameservers 192.168.2.108 bridge_ports eth0 bridge_fd 9 bridge_hello 2 bridge_maxage 12 bridge_maxwait 5 bridge_stp off ok? # define the virtual network vm_lan.template: NAME = VM LAN TYPE = RANGED BRIDGE = br0 NETWORK_ADDRESS = 192.168.2.128 NETWORK_SIZE = 126 NETMASK = 255.255.255.0 GATEWAY = 192.168.2.108 NS = 192.168.2.108 # ip of frontend ok? # define xen settings /srv/cloud/one/etc/vmm_sh/vmm_sh_xen.conf CPU = 0.5 MEMORY = 256 OS = [ KERNEL=/boot/vmlinuz-2.6.26-1-xen-amd64, INITRD=/boot/initrd.img-2.6.26-1-xen-amd64 , kernel_cmd=ro, root=sda5 ] CREDIT = 256 ok? # check etc/oned.conf VM_DIR=/srv/cloud/one/var/images IMAGE_REPOSITORY_PATH = /srv/cloud/one/var/images DEFAULT_IMAGE_TYPE = CDROM # because of iso-image, is this ok? DEFAULT_DEVICE_PREFIX = hd VM_MAD = [ name = vmm_xen, executable = one_vmm_sh, arguments = xen, default = vmm_sh/vmm_sh_xen.conf, type = xen ] [...] ok? # check vm_instance.template NAME = debian CPU = 1 MEMORY = 512 # Define your main disk as usual, and tell OpenNebula to keep the changes. DISK = [ TYPE = disk, SOURCE = /srv/cloud/images/debian_disk.img, TARGET = hda, CLONE = no ] DISK = [ TYPE = cdrom, SOURCE = /srv/cloud/images/debian-506-amd64-netinst.iso, TARGET = hdb, CLONE = no ] # swap, sdd DISK = [ TYPE = swap, SIZE = 1024, READONLY = no ] DISK = [ TYPE = fs, SIZE = 4096, FORMAT = ext3, SAVE = yes, TARGET = sdg ] NIC = [ NETWORK = VM LAN ] GRAPHICS = [ type = vnc, listen = node1, port = 5902 ] FEATURES=[ acpi=no ] vielen dank nochmal gruß sufian Exklusiv: Neue E-Mail-Adresse @iPhone.de jetzt verfügbar! Sichern Sie sich jetzt ihre persönliche e-mail-adre...@iphone.de___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] Virtual machine IP
Hi Steven, Thanks a lot for your reply. Do you have the latest one-context script from opennebula 2.0 included as part of your VM.. that is what is either supposed to grab the IP that opennebula assigns through contextualization, or make it be the last 4 bytes of the MAC address. At least that's what I understand. Yes I think I have included in the VM. Is your image configured to get DHCP network otherwise? No I don't have DHCP but since I had the one-context scripts, I didn't think it was necessary. Better to create a different bridge from virbr0 especially for opennebula networking.. if you are doing KVM you will have to do this by hand with brctl. I'm using the default bridge created by libvirt, and I agree that it is probably where the issue is coming from. How would I create a new bridge for OpenNebula to use? I have already tried creating one but didn't succeed... so I used the vibr0 instead since it was already there. Just to give you an idea of how my network looks like: The frontend and two nodes are connected to eth0. I would appreciate some advice on how to accomplish this: I think I'm going to run these commands on all machines. Then I will just indicate br0 in the VNet template for OpenNebula. brctl addbr br0 brctl addif br0 eth0 ifconfig br0 up Is this all I should do? Thank you Zaina ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
[one-users] New VirtualBox Driver
Hi, These days I've been writing some scripts to make OpenNebula 2.0 work with virtualbox, since I need something simpler than xen, and the machines I use doesn't have kvm support. It is a little limited for now, but can do the basic things and I believe it can be useful to somebody. It was written in Bash and Python. Here is the address: http://code.google.com/p/vbox-one Any feedback will be appreciated. []s Henrique Teófilo ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
Re: [one-users] Virtual machine IP
Hi, In [1] you have a useful link to create bridges. I also had problems to get up the interfaces with the default vmcontext.sh. So, I added at the end ifup -a. [1] http://wiki.libvirt.org/page/Networking#Bridged_networking_.28aka_.22shared_physical_device.22.29 Regards, Javi El 08/12/2010 15:15, zaina.afoulki escribió: Hi Steven, Thanks a lot for your reply. Do you have the latest one-context script from opennebula 2.0 included as part of your VM.. that is what is either supposed to grab the IP that opennebula assigns through contextualization, or make it be the last 4 bytes of the MAC address. At least that's what I understand. Yes I think I have included in the VM. Is your image configured to get DHCP network otherwise? No I don't have DHCP but since I had the one-context scripts, I didn't think it was necessary. Better to create a different bridge from virbr0 especially for opennebula networking.. if you are doing KVM you will have to do this by hand with brctl. I'm using the default bridge created by libvirt, and I agree that it is probably where the issue is coming from. How would I create a new bridge for OpenNebula to use? I have already tried creating one but didn't succeed... so I used the vibr0 instead since it was already there. Just to give you an idea of how my network looks like: The frontend and two nodes are connected to eth0. I would appreciate some advice on how to accomplish this: I think I'm going to run these commands on all machines. Then I will just indicate br0 in the VNet template for OpenNebula. brctl addbr br0 brctl addif br0 eth0 ifconfig br0 up Is this all I should do? Thank you Zaina ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org -- +-+ Dr. Javier Diaz Post Doctoral Fellow Pervasive Technology Institute Indiana University 2719 East Tenth Street, Bloomington, IN 47408, USA e-mail: javid...@indiana.edu +-+ ___ Users mailing list Users@lists.opennebula.org http://lists.opennebula.org/listinfo.cgi/users-opennebula.org