Re: [VCL 2.2.1] [Power7] Problem with image reservation
Aaron, I wanted to try the ssh method to change the root password, but I stopped short since it will ask for old password before changing to a new one. However, I planning to create and make use of user accounts instead of looking for the root password. Thank you for the input. Regards, Sunil On Aug 4, 2011, at 12:27 PM, Aaron Peeler wrote: Sunil, On #2. The VCL load process is to randomize the root (and administrator for windows) password. It should be in the vcld.log file, but an easier option might be to ssh into the node and set the root password to something you know when using xcat's rcons to look at the console. Aaron On Tue, Aug 2, 2011 at 2:17 PM, Sunil Venkatesh suni...@umbc.edu wrote: Hi Josh, So, I was able to get the VCL to capture restore the images on to the Power7 blade without any errors. The VCL web portal shows the Power7 node being loaded with the captured image. I was looking for some clarification with a couple of things. 1. The web portal shows Selection currently not available when I intend to make a reservation on the Power7 blade that I have been working with all this while. 2. Once the image is loaded onto the blade, I am able to login from the management node without a password. However, when I am using rcons to login as root on the Power7 blade, it does not accept the root password I had used during the OS installation. Does the root password get reset to a default one? I was checking vcl/lib/VCL/Module/OS/Linux.pm if that is the case. Thanks Regards, Sunil On Jul 7, 2011, at 2:03 PM, Josh Thompson wrote: This and your next question are both deeper into the backend code that I've worked with. Andy or Aaron may be able to answer your questions further. Josh -- Aaron Peeler Program Manager Virtual Computing Lab NC State University All electronic mail messages in connection with State business which are sent to or received by this account are subject to the NC Public Records Law and may be disclosed to third parties.
Re: [VCL 2.2.1] [Power7] Problem with image reservation
interesting it's prompting, which version of linux. The code is using: echo $passwd | /usr/bin/passwd -f $account --stdin sounds this it worked in the vcl code, so should also work at the cmdline. Aaron On Thu, Aug 4, 2011 at 12:49 PM, Sunil Venkatesh suni...@umbc.edu wrote: Aaron, I wanted to try the ssh method to change the root password, but I stopped short since it will ask for old password before changing to a new one. However, I planning to create and make use of user accounts instead of looking for the root password. Thank you for the input. Regards, Sunil On Aug 4, 2011, at 12:27 PM, Aaron Peeler wrote: Sunil, On #2. The VCL load process is to randomize the root (and administrator for windows) password. It should be in the vcld.log file, but an easier option might be to ssh into the node and set the root password to something you know when using xcat's rcons to look at the console. Aaron On Tue, Aug 2, 2011 at 2:17 PM, Sunil Venkatesh suni...@umbc.edu wrote: Hi Josh, So, I was able to get the VCL to capture restore the images on to the Power7 blade without any errors. The VCL web portal shows the Power7 node being loaded with the captured image. I was looking for some clarification with a couple of things. 1. The web portal shows Selection currently not available when I intend to make a reservation on the Power7 blade that I have been working with all this while. 2. Once the image is loaded onto the blade, I am able to login from the management node without a password. However, when I am using rcons to login as root on the Power7 blade, it does not accept the root password I had used during the OS installation. Does the root password get reset to a default one? I was checking vcl/lib/VCL/Module/OS/Linux.pm if that is the case. Thanks Regards, Sunil On Jul 7, 2011, at 2:03 PM, Josh Thompson wrote: This and your next question are both deeper into the backend code that I've worked with. Andy or Aaron may be able to answer your questions further. Josh -- Aaron Peeler Program Manager Virtual Computing Lab NC State University All electronic mail messages in connection with State business which are sent to or received by this account are subject to the NC Public Records Law and may be disclosed to third parties. -- Aaron Peeler Program Manager Virtual Computing Lab NC State University All electronic mail messages in connection with State business which are sent to or received by this account are subject to the NC Public Records Law and may be disclosed to third parties.
Re: [VCL 2.2.1] [Power7] Problem with image reservation
Josh, Could you provide me with the links to the resources of the VCL workshop? If there is a way to witness the workshop while it is in progress, that would help too. Regards, Sunil On Jul 7, 2011, at 2:03 PM, Josh Thompson wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Sunil, On Thursday July 07, 2011, Sunil Venkatesh wrote: Thanks Josh. My professor was asking about the details of VCL workshop in NC. Are you aware of these details? The workshop is hosted by NCSU. It takes people from an introduction to VCL to actually installing and managing it. It is already full, but I think recordings of the sessions may be available when it is over. Please bare with my comments inline. Responses also inline. On 7/7/11 11:13 AM, Josh Thompson wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Tuesday July 05, 2011, Sunil Venkatesh wrote: Hi Josh, I was able to get the following things done in respect to getting VCL to work on POWER. 1. Made modifications in the xcat tables to get the capture process working with statelite images instead of stateless images. Particularly the noderes bootparams table. 2. Used partimage to capture the images (did NOT set usepartimageng to 1). -rw-r--r-- 1 root root0 Jul 5 16:38 compute.img.capturedone -rw-r--r-- 1 root root0 Jul 5 15:58 compute.img.capturefailed -rw--- 1 root root 6.5M Jul 5 16:07 compute-parta2.gz -rw--- 1 root root 679M Jul 5 16:10 compute-parta3.gz -rw--- 1 root root 23M Jul 5 16:38 compute-parta6.gz -rw-r--r-- 1 root root 512 Jul 5 16:07 compute-sda.mbr -rw-r--r-- 1 root root 363 Jul 5 16:07 compute-sda.sfdisk 2 partitions including the boot partition present on the blade were captured under /install/image/ppc64/. Initially, RHEL 5 was installed on a 600 GB partition due to which the capture process failed. The image of the partition was generated once the partition size was reduced to 6GB. Is it necessary for me to use partimage-ng instead of partimage itself? Are you asking if you need to use partimage-ng for partitions that are 600GB? If so, I don't really know. We've never dealt with partitions that large. Here, I am just asking if images captured using partimage are recognized by VCL or is it required that I use partimage-ng. From your earlier emails to Prem, I could notice that the only difference between partimage partimage-ng (after setting userpartimageng to 1) is the former generates images with .gz and the later generates .img. Am I right here? Also, I was able to get the 600GB partition captured, since the partition was empty, it resulted in a ~17MB image file. VCL can deploy images captured with both partimage and partimage-ng. At NCSU, we were going to switch to partimage-ng, which is why I added in support for it, but then we realized we'd have to upgrade all of our management nodes to xCAT2 at the same time or some of them wouldn't be able to deploy newly captured images that were captured with partimage-ng (the support for xCAT1.x can't deploy using partimage-ng). So, we just stuck with partimage. The captured file format between the two is different. When proceeding further with vcld --setup, the script was not able to find the images that were created using partimage. The options that are provided in the script does not allow for selecting an architecture other than x86/x86_64. You'll need to modify the vcld image.pm module. Look in /usr/local/vcl/lib/VCL. In image.pm, look for the function 'setup_capture_base_image'; then, find 'my @architecture_choices' and add 'ppc' as another option. As a matter of fact, I tried this step. But, the _get_image_repository_path function in /usr/local/vcl/lib/VCL/Module/Provisioning/xCAT.pm does not recognize the architecture when I choose ppc/ppc64 in the menu. On line 2922 in the same file, image_architecture is set to undefined. I think the list of supported architectures is stored in some mysql table. I haven't checked regarding this, i was trying to get VCL to recognize the images as x86/x86_64 by setting up soft links in the search paths of VCL. This and your next question are both deeper into the backend code that I've worked with. Andy or Aaron may be able to answer your questions further. Josh Also, in the error log vcld is looking for /opt/xcat/share/xcat/install/image/rh5image-power010701bi34-v0.tmpl and cannot find the template file. Should the template file that needs to be accessed in this case be createimage.ppc64.tmpl? This is actually a check to make sure the image doesn't already exist before trying to capturing it. So, it is good that it doesn't find it. If possible, could you please provide me with the details of steps that take place here. If there are any documentation available regarding this, that would work too. U said image doesn't already exist before trying to capturing it, how
Re: [VCL 2.2.1] [Power7] Problem with image reservation
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Tuesday July 05, 2011, Sunil Venkatesh wrote: Hi Josh, I was able to get the following things done in respect to getting VCL to work on POWER. 1. Made modifications in the xcat tables to get the capture process working with statelite images instead of stateless images. Particularly the noderes bootparams table. 2. Used partimage to capture the images (did NOT set usepartimageng to 1). -rw-r--r-- 1 root root0 Jul 5 16:38 compute.img.capturedone -rw-r--r-- 1 root root0 Jul 5 15:58 compute.img.capturefailed -rw--- 1 root root 6.5M Jul 5 16:07 compute-parta2.gz -rw--- 1 root root 679M Jul 5 16:10 compute-parta3.gz -rw--- 1 root root 23M Jul 5 16:38 compute-parta6.gz -rw-r--r-- 1 root root 512 Jul 5 16:07 compute-sda.mbr -rw-r--r-- 1 root root 363 Jul 5 16:07 compute-sda.sfdisk 2 partitions including the boot partition present on the blade were captured under /install/image/ppc64/. Initially, RHEL 5 was installed on a 600 GB partition due to which the capture process failed. The image of the partition was generated once the partition size was reduced to 6GB. Is it necessary for me to use partimage-ng instead of partimage itself? Are you asking if you need to use partimage-ng for partitions that are 600GB? If so, I don't really know. We've never dealt with partitions that large. When proceeding further with vcld --setup, the script was not able to find the images that were created using partimage. The options that are provided in the script does not allow for selecting an architecture other than x86/x86_64. You'll need to modify the vcld image.pm module. Look in /usr/local/vcl/lib/VCL. In image.pm, look for the function 'setup_capture_base_image'; then, find 'my @architecture_choices' and add 'ppc' as another option. Also, in the error log vcld is looking for /opt/xcat/share/xcat/install/image/rh5image-power010701bi34-v0.tmpl and cannot find the template file. Should the template file that needs to be accessed in this case be createimage.ppc64.tmpl? This is actually a check to make sure the image doesn't already exist before trying to capturing it. So, it is good that it doesn't find it. It sounds like you're almost there. Great work! Josh I have attached a log at the end of the mail. I am not sure where I have gone wrong with the VCL configuration. -Sunil - rh5image-power010701bi34-v0 image creation failed time: 2011-07-05 11:03:25 caller: image.pm:reservation_failed(385) ( 0) image.pm, reservation_failed (line: 385) (-1) image.pm, process (line: 167) (-2) vcld, make_new_child (line: 568) (-3) vcld, main (line: 346) management node: web1.bluegrit.cs.umbc.edu reservation PID: 9866 parent vcld PID: 19110 request ID: 30 reservation ID: 30 request state/laststate: image/image request start time: 2011-07-05 11:03:20 request end time: 2011-07-05 12:03:20 for imaging: no log ID: none computer: power01.bluegrit.cs.umbc.edu computer id: 2 computer type: blade computer eth0 MAC address:undefined computer eth1 MAC address:undefined computer private IP address: 172.20.106.1 computer public IP address: 172.20.106.1 computer in block allocation: no provisioning module: VCL::Module::Provisioning::xCAT2 image: rh5image-power010701bi34-v0 image display name: power010701bi image ID: 34 image revision ID: 34 image size: 1450 MB use Sysprep: yes root access: yes image owner ID: 1 image owner affiliation: Local image revision date created: 2011-07-05 11:03:25 image revision production: yes OS module: VCL::Module::OS::Linux user: admin user name: vcl admin user ID: 1 user affiliation: Local RECENT LOG ENTRIES FOR THIS PROCESS: 2011-07-05 11:03:25|9866|30:30|image|Module.pm:create_os_object(304)|VCL::Module::OS: :Linux OS object created for rh5image-power010701bi34-v0, address: 88fb070 2011-07-05 11:03:25|9866|30:30|image|xCAT.pm:initialize(110)|XCATROOT environment variable is not set, using /opt/xcat 2011-07-05 11:03:25|9866|30:30|image|xCAT.pm:initialize(128)|xCAT root path found: /opt/xcat 2011-07-05 11:03:25|9866|30:30|image|xCAT.pm:initialize(130)|xCAT module initialized 2011-07-05 11:03:25|9866|30:30|image|xCAT2.pm:initialize(110)|XCATROOT environment variable is not set, using /opt/xcat 2011-07-05 11:03:25|9866|30:30|image|xCAT2.pm:initialize(128)|xCAT root path found: /opt/xcat 2011-07-05 11:03:25|9866|30:30|image|xCAT2.pm:initialize(130)|xCAT module initialized 2011-07-05 11:03:25|9866|30:30|image|Module.pm:create_provisioning_object(420)|VCL::M odule::Provisioning::xCAT2 module loaded 2011-07-05 11:03:25|9866|30:30|image|Module.pm:create_mn_os_object(335)|management node OS object
Re: [VCL 2.2.1] [Power7] Problem with image reservation
Thanks Josh. My professor was asking about the details of VCL workshop in NC. Are you aware of these details? Please bare with my comments inline. On 7/7/11 11:13 AM, Josh Thompson wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Tuesday July 05, 2011, Sunil Venkatesh wrote: Hi Josh, I was able to get the following things done in respect to getting VCL to work on POWER. 1. Made modifications in the xcat tables to get the capture process working with statelite images instead of stateless images. Particularly the noderes bootparams table. 2. Used partimage to capture the images (did NOT set usepartimageng to 1). -rw-r--r-- 1 root root0 Jul 5 16:38 compute.img.capturedone -rw-r--r-- 1 root root0 Jul 5 15:58 compute.img.capturefailed -rw--- 1 root root 6.5M Jul 5 16:07 compute-parta2.gz -rw--- 1 root root 679M Jul 5 16:10 compute-parta3.gz -rw--- 1 root root 23M Jul 5 16:38 compute-parta6.gz -rw-r--r-- 1 root root 512 Jul 5 16:07 compute-sda.mbr -rw-r--r-- 1 root root 363 Jul 5 16:07 compute-sda.sfdisk 2 partitions including the boot partition present on the blade were captured under /install/image/ppc64/. Initially, RHEL 5 was installed on a 600 GB partition due to which the capture process failed. The image of the partition was generated once the partition size was reduced to 6GB. Is it necessary for me to use partimage-ng instead of partimage itself? Are you asking if you need to use partimage-ng for partitions that are 600GB? If so, I don't really know. We've never dealt with partitions that large. Here, I am just asking if images captured using partimage are recognized by VCL or is it required that I use partimage-ng. From your earlier emails to Prem, I could notice that the only difference between partimage partimage-ng (after setting userpartimageng to 1) is the former generates images with .gz and the later generates .img. Am I right here? Also, I was able to get the 600GB partition captured, since the partition was empty, it resulted in a ~17MB image file. When proceeding further with vcld --setup, the script was not able to find the images that were created using partimage. The options that are provided in the script does not allow for selecting an architecture other than x86/x86_64. You'll need to modify the vcld image.pm module. Look in /usr/local/vcl/lib/VCL. In image.pm, look for the function 'setup_capture_base_image'; then, find 'my @architecture_choices' and add 'ppc' as another option. As a matter of fact, I tried this step. But, the _get_image_repository_path function in /usr/local/vcl/lib/VCL/Module/Provisioning/xCAT.pm does not recognize the architecture when I choose ppc/ppc64 in the menu. On line 2922 in the same file, image_architecture is set to undefined. I think the list of supported architectures is stored in some mysql table. I haven't checked regarding this, i was trying to get VCL to recognize the images as x86/x86_64 by setting up soft links in the search paths of VCL. Also, in the error log vcld is looking for /opt/xcat/share/xcat/install/image/rh5image-power010701bi34-v0.tmpl and cannot find the template file. Should the template file that needs to be accessed in this case be createimage.ppc64.tmpl? This is actually a check to make sure the image doesn't already exist before trying to capturing it. So, it is good that it doesn't find it. If possible, could you please provide me with the details of steps that take place here. If there are any documentation available regarding this, that would work too. U said image doesn't already exist before trying to capturing it, how does VCL capture the images? does it make use of the images that are already generated using partimage? if so, in what places does it look for the images? Sorry for asking too many questions. I could trace the scripts to check the flow, but, that would take a lot of time. You have been really patient with all my queries, appreciate that. Thanks Sunil It sounds like you're almost there. Great work! Josh I have attached a log at the end of the mail. I am not sure where I have gone wrong with the VCL configuration. -Sunil - rh5image-power010701bi34-v0 image creation failed time: 2011-07-05 11:03:25 caller: image.pm:reservation_failed(385) ( 0) image.pm, reservation_failed (line: 385) (-1) image.pm, process (line: 167) (-2) vcld, make_new_child (line: 568) (-3) vcld, main (line: 346) management node: web1.bluegrit.cs.umbc.edu reservation PID: 9866 parent vcld PID: 19110 request ID: 30 reservation ID: 30 request state/laststate: image/image request start time: 2011-07-05 11:03:20 request end time: 2011-07-05 12:03:20 for imaging: no log ID: none computer: power01.bluegrit.cs.umbc.edu computer id: 2 computer type: blade computer eth0 MAC address:undefined computer eth1
Re: [VCL 2.2.1] [Power7] Problem with image reservation
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Sunil, On Thursday July 07, 2011, Sunil Venkatesh wrote: Thanks Josh. My professor was asking about the details of VCL workshop in NC. Are you aware of these details? The workshop is hosted by NCSU. It takes people from an introduction to VCL to actually installing and managing it. It is already full, but I think recordings of the sessions may be available when it is over. Please bare with my comments inline. Responses also inline. On 7/7/11 11:13 AM, Josh Thompson wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Tuesday July 05, 2011, Sunil Venkatesh wrote: Hi Josh, I was able to get the following things done in respect to getting VCL to work on POWER. 1. Made modifications in the xcat tables to get the capture process working with statelite images instead of stateless images. Particularly the noderes bootparams table. 2. Used partimage to capture the images (did NOT set usepartimageng to 1). -rw-r--r-- 1 root root0 Jul 5 16:38 compute.img.capturedone -rw-r--r-- 1 root root0 Jul 5 15:58 compute.img.capturefailed -rw--- 1 root root 6.5M Jul 5 16:07 compute-parta2.gz -rw--- 1 root root 679M Jul 5 16:10 compute-parta3.gz -rw--- 1 root root 23M Jul 5 16:38 compute-parta6.gz -rw-r--r-- 1 root root 512 Jul 5 16:07 compute-sda.mbr -rw-r--r-- 1 root root 363 Jul 5 16:07 compute-sda.sfdisk 2 partitions including the boot partition present on the blade were captured under /install/image/ppc64/. Initially, RHEL 5 was installed on a 600 GB partition due to which the capture process failed. The image of the partition was generated once the partition size was reduced to 6GB. Is it necessary for me to use partimage-ng instead of partimage itself? Are you asking if you need to use partimage-ng for partitions that are 600GB? If so, I don't really know. We've never dealt with partitions that large. Here, I am just asking if images captured using partimage are recognized by VCL or is it required that I use partimage-ng. From your earlier emails to Prem, I could notice that the only difference between partimage partimage-ng (after setting userpartimageng to 1) is the former generates images with .gz and the later generates .img. Am I right here? Also, I was able to get the 600GB partition captured, since the partition was empty, it resulted in a ~17MB image file. VCL can deploy images captured with both partimage and partimage-ng. At NCSU, we were going to switch to partimage-ng, which is why I added in support for it, but then we realized we'd have to upgrade all of our management nodes to xCAT2 at the same time or some of them wouldn't be able to deploy newly captured images that were captured with partimage-ng (the support for xCAT1.x can't deploy using partimage-ng). So, we just stuck with partimage. The captured file format between the two is different. When proceeding further with vcld --setup, the script was not able to find the images that were created using partimage. The options that are provided in the script does not allow for selecting an architecture other than x86/x86_64. You'll need to modify the vcld image.pm module. Look in /usr/local/vcl/lib/VCL. In image.pm, look for the function 'setup_capture_base_image'; then, find 'my @architecture_choices' and add 'ppc' as another option. As a matter of fact, I tried this step. But, the _get_image_repository_path function in /usr/local/vcl/lib/VCL/Module/Provisioning/xCAT.pm does not recognize the architecture when I choose ppc/ppc64 in the menu. On line 2922 in the same file, image_architecture is set to undefined. I think the list of supported architectures is stored in some mysql table. I haven't checked regarding this, i was trying to get VCL to recognize the images as x86/x86_64 by setting up soft links in the search paths of VCL. This and your next question are both deeper into the backend code that I've worked with. Andy or Aaron may be able to answer your questions further. Josh Also, in the error log vcld is looking for /opt/xcat/share/xcat/install/image/rh5image-power010701bi34-v0.tmpl and cannot find the template file. Should the template file that needs to be accessed in this case be createimage.ppc64.tmpl? This is actually a check to make sure the image doesn't already exist before trying to capturing it. So, it is good that it doesn't find it. If possible, could you please provide me with the details of steps that take place here. If there are any documentation available regarding this, that would work too. U said image doesn't already exist before trying to capturing it, how does VCL capture the images? does it make use of the images that are already generated using partimage? if so, in what places does it look for the images? Sorry for asking too many questions. I could trace the scripts
Re: [VCL 2.2.1] [Power7] Problem with image reservation
Hi, Update ! I was able to fix the problem that I was facing with the scripts by disabling the firewall. But, I still have a problem with the command- nodeset nodename image Unless this error is fixed, I don't think partimage will work. Am I right here? Thanks, Sunil On Tue, Jun 21, 2011 at 3:13 PM, Sunil Venkatesh suni...@umbc.edu wrote: Josh, I have reached a point where I am able to boot the ppc using the statelite images created using genimage. But, I was wondering how significant the following command is. nodeset nodename image I got the same error that Prem had mentioned. power01: Error: Unable to identify plugin for this command, check relevant tables: nodetype.os Error: Some nodes failed to set up image resources, aborting I tried changing the 'os' field to 'image' under nodetype, that doesn't seem to help. I get the same error even after the change. 'arch' in my case is set to 'ppc64'. Also, I think partimage plugin needs to be changed to support the ppc architecture, from what you had mentioned in the other thread. I am not sure what the command 'nodeset nodename image' does, but, I am able to boot the statelite images by making changes to the yaboot configuration files. The ppc blade currently uses LVM, that needs to be replaced with ext2/ext3 from what I read from the other thread, am I right? Also, just out of curiosity I left the statelite image to boot with my current setting. I can see the xcat script throwing an error- /opt/xcat/xcatdsklspost: line 229: /xcatpost/getpostscript.awk: No such file or directory /tmp/mypostscript: line 16: updateflag.awk: command not found both getpostscript.awk updateflag.awk are not found in the rootimg created by genimage. Is there any place I could find these scripts? Also, please correct me if there is anything wrong with the procedure I am following. Thanks in advance. Regards, Sunil On 6/13/11 4:13 PM, Josh Thompson wrote: Sunil, From what I remember, I didn't have to do much to the rootimg.gz image to make it work. I created the files I supply before xCAT started using statelite instead of stateless. I think statelite uses NFS to mount the image, and stateless uses an image file downloaded to the node and run out of RAM. Since generating a statelite image is pretty straightforward use of xCAT, you may want to ask on the xcat-user email list for help with it. Unless you can have the admins of the other dhcp server on your network exclude the MAC addresses of your blades, you'll need to create a separate private network to control your VCL stuff, either physically or with VLANs. If they can exclude the MACs, you can set up the dhcp server on your management node to only answer to requests from your blades. Josh On Monday June 13, 2011, Sunil Venkatesh wrote: Josh, Again, Thank you for your valuable inputs. I have got to the point where I can get the compute node to boot using the stateless images. I had to manually configure the netboot since we already had a DHCP server which is not the same as our Management node. Since our setup is not in an isolated environment, I could not let xcat handle the dhcp netboot configuration (it messed up out network configuration when i let xcat handle it,we had 2 dhcp servers running at that point). Are you aware of any way to let xcat handle such scenarios? Although I am able to get the compute node to boot with the kernel image initrd, and NFS mount the rootimg that was generated using 'genimage', I am getting the following error on the compute node's console - FATAL error: could not get the entries from litefile table... after going thru the init-scripts, I found out 'xCATCmd' binary is not present in the rootimg. I am currently checking the xcat packages for its availability. If you know the procedure to get it onto the compute node, please let me know the same. Appreciate your support. Thanking you, Sunil On 6/8/11 9:02 AM, Josh Thompson wrote: Sunil, I don't recall seeing any documentation on those parts. I had to poke around looking at parts of xCAT to see how it worked. It's been a few years since I did that; so, I don't remember much about the process. My recommendation would be to start looking at things in the rootimg.gz image. Looking at it now, I see that /opt/xcat/xcatdsklspost gets run when rootimg.gz boots. It looks like it downloads all of the postscripts from the management node and then run getpostscript.awk which issues a command to xcatd to get the primary postscript for that machine. I've forgotten how xcatd then builds the primary postscript. I do remember that in the partimageng.pm module, I had it add the partimageng postscript. So, you'll really have to start digging through how the xcat postscript system works. Josh On Tuesday June 07, 2011, Sunil Venkatesh wrote: Josh, Is there any place I could find some details on ... /Once the compute
Re: [VCL 2.2.1] [Power7] Problem with image reservation
Josh, I have reached a point where I am able to boot the ppc using the statelite images created using genimage. But, I was wondering how significant the following command is. nodeset nodename image I got the same error that Prem had mentioned. power01: Error: Unable to identify plugin for this command, check relevant tables: nodetype.os Error: Some nodes failed to set up image resources, aborting I tried changing the 'os' field to 'image' under nodetype, that doesn't seem to help. I get the same error even after the change. 'arch' in my case is set to 'ppc64'. Also, I think partimage plugin needs to be changed to support the ppc architecture, from what you had mentioned in the other thread. I am not sure what the command 'nodeset nodename image' does, but, I am able to boot the statelite images by making changes to the yaboot configuration files. The ppc blade currently uses LVM, that needs to be replaced with ext2/ext3 from what I read from the other thread, am I right? Also, just out of curiosity I left the statelite image to boot with my current setting. I can see the xcat script throwing an error- /opt/xcat/xcatdsklspost: line 229: /xcatpost/getpostscript.awk: No such file or directory /tmp/mypostscript: line 16: updateflag.awk: command not found both getpostscript.awk updateflag.awk are not found in the rootimg created by genimage. Is there any place I could find these scripts? Also, please correct me if there is anything wrong with the procedure I am following. Thanks in advance. Regards, Sunil On 6/13/11 4:13 PM, Josh Thompson wrote: Sunil, From what I remember, I didn't have to do much to the rootimg.gz image to make it work. I created the files I supply before xCAT started using statelite instead of stateless. I think statelite uses NFS to mount the image, and stateless uses an image file downloaded to the node and run out of RAM. Since generating a statelite image is pretty straightforward use of xCAT, you may want to ask on the xcat-user email list for help with it. Unless you can have the admins of the other dhcp server on your network exclude the MAC addresses of your blades, you'll need to create a separate private network to control your VCL stuff, either physically or with VLANs. If they can exclude the MACs, you can set up the dhcp server on your management node to only answer to requests from your blades. Josh On Monday June 13, 2011, Sunil Venkatesh wrote: Josh, Again, Thank you for your valuable inputs. I have got to the point where I can get the compute node to boot using the stateless images. I had to manually configure the netboot since we already had a DHCP server which is not the same as our Management node. Since our setup is not in an isolated environment, I could not let xcat handle the dhcp netboot configuration (it messed up out network configuration when i let xcat handle it,we had 2 dhcp servers running at that point). Are you aware of any way to let xcat handle such scenarios? Although I am able to get the compute node to boot with the kernel image initrd, and NFS mount the rootimg that was generated using 'genimage', I am getting the following error on the compute node's console - FATAL error: could not get the entries from litefile table... after going thru the init-scripts, I found out 'xCATCmd' binary is not present in the rootimg. I am currently checking the xcat packages for its availability. If you know the procedure to get it onto the compute node, please let me know the same. Appreciate your support. Thanking you, Sunil On 6/8/11 9:02 AM, Josh Thompson wrote: Sunil, I don't recall seeing any documentation on those parts. I had to poke around looking at parts of xCAT to see how it worked. It's been a few years since I did that; so, I don't remember much about the process. My recommendation would be to start looking at things in the rootimg.gz image. Looking at it now, I see that /opt/xcat/xcatdsklspost gets run when rootimg.gz boots. It looks like it downloads all of the postscripts from the management node and then run getpostscript.awk which issues a command to xcatd to get the primary postscript for that machine. I've forgotten how xcatd then builds the primary postscript. I do remember that in the partimageng.pm module, I had it add the partimageng postscript. So, you'll really have to start digging through how the xcat postscript system works. Josh On Tuesday June 07, 2011, Sunil Venkatesh wrote: Josh, Is there any place I could find some details on ... /Once the compute node is booted with the stateless image, it uses NFS to mount some things from the management node, and then runs some xcat postscripts,/ I have the stateless images ready with partimage compiled for PPC. For the compute node (power 7) to boot using the stateless images, i need to configure the yaboot instead of pxeboot (which is specific to x86). I wanted to know where in the startup files the execution of
Re: [VCL 2.2.1] [Power7] Problem with image reservation
Josh, Again, Thank you for your valuable inputs. I have got to the point where I can get the compute node to boot using the stateless images. I had to manually configure the netboot since we already had a DHCP server which is not the same as our Management node. Since our setup is not in an isolated environment, I could not let xcat handle the dhcp netboot configuration (it messed up out network configuration when i let xcat handle it,we had 2 dhcp servers running at that point). Are you aware of any way to let xcat handle such scenarios? Although I am able to get the compute node to boot with the kernel image initrd, and NFS mount the rootimg that was generated using 'genimage', I am getting the following error on the compute node's console - FATAL error: could not get the entries from litefile table... after going thru the init-scripts, I found out 'xCATCmd' binary is not present in the rootimg. I am currently checking the xcat packages for its availability. If you know the procedure to get it onto the compute node, please let me know the same. Appreciate your support. Thanking you, Sunil On 6/8/11 9:02 AM, Josh Thompson wrote: Sunil, I don't recall seeing any documentation on those parts. I had to poke around looking at parts of xCAT to see how it worked. It's been a few years since I did that; so, I don't remember much about the process. My recommendation would be to start looking at things in the rootimg.gz image. Looking at it now, I see that /opt/xcat/xcatdsklspost gets run when rootimg.gz boots. It looks like it downloads all of the postscripts from the management node and then run getpostscript.awk which issues a command to xcatd to get the primary postscript for that machine. I've forgotten how xcatd then builds the primary postscript. I do remember that in the partimageng.pm module, I had it add the partimageng postscript. So, you'll really have to start digging through how the xcat postscript system works. Josh On Tuesday June 07, 2011, Sunil Venkatesh wrote: Josh, Is there any place I could find some details on ... /Once the compute node is booted with the stateless image, it uses NFS to mount some things from the management node, and then runs some xcat postscripts,/ I have the stateless images ready with partimage compiled for PPC. For the compute node (power 7) to boot using the stateless images, i need to configure the yaboot instead of pxeboot (which is specific to x86). I wanted to know where in the startup files the execution of partimage and NFS mount is configured. Is it configured by the genimage command itself? Considering the way in which the nodes are configured in the network, it would not be a good idea to let xcat take care of configuring the details like DHCPD for netboot. So, I need to make changes to the configuration files manually, which is why this query came up. Thanks in advance. Regards, Sunil On 6/1/11 1:39 PM, Josh Thompson wrote: Sunil, The stateless image I refer to is what is actually booted on the compute node containing the image to be captured. It's called stateless because it is loaded completely in RAM and does not maintain any state when a reboot occurs. The partimage binary is part of this stateless image and actually runs on the compute node. It does not run on the management node. The management node does not have block level access to the disk on the compute node to be able to capture the image from the disk. I'll try to describe the process a little better. The management node issues a reboot command to the compute node. The compute node uses PXE to load and boot a kernel (vmlinuz), initial RAM disk (initrd.img), and a root filesystem (rootimg.gz) from the management node. All three of these together make up the stateless image. Once the compute node is booted with the stateless image, it uses NFS to mount some things from the management node, and then runs some xcat postscripts, one of which is the partimageng postscript. This postscript determines what partitions are on the compute node and, depending on how the postscript is configured, uses partimage or partimageng to capture an image of the compute node disk that is then saved to the management node. When it is finished capturing the image, it notifies xcat on the management node and then reboots. xcat reconfigures itself to tell the compute node to boot off of disk at next boot. When the compute node comes up, it uses PXE to ask the management node how to boot. The management node tells it to boot off of disk. I hope that clarifies how the system works. If any of it is unclear, please ask for further clarification. Josh On Wednesday June 01, 2011, Sunil Venkatesh wrote: Josh, I had one more clarification. partimage binaries run in the management node to capture an (stateless) image from the compute node right? In that case, is there a need for these binaries to go into the rootimg.gz?? My assumption is, partimage runs on
Re: [VCL 2.2.1] [Power7] Problem with image reservation
Sunil, From what I remember, I didn't have to do much to the rootimg.gz image to make it work. I created the files I supply before xCAT started using statelite instead of stateless. I think statelite uses NFS to mount the image, and stateless uses an image file downloaded to the node and run out of RAM. Since generating a statelite image is pretty straightforward use of xCAT, you may want to ask on the xcat-user email list for help with it. Unless you can have the admins of the other dhcp server on your network exclude the MAC addresses of your blades, you'll need to create a separate private network to control your VCL stuff, either physically or with VLANs. If they can exclude the MACs, you can set up the dhcp server on your management node to only answer to requests from your blades. Josh On Monday June 13, 2011, Sunil Venkatesh wrote: Josh, Again, Thank you for your valuable inputs. I have got to the point where I can get the compute node to boot using the stateless images. I had to manually configure the netboot since we already had a DHCP server which is not the same as our Management node. Since our setup is not in an isolated environment, I could not let xcat handle the dhcp netboot configuration (it messed up out network configuration when i let xcat handle it,we had 2 dhcp servers running at that point). Are you aware of any way to let xcat handle such scenarios? Although I am able to get the compute node to boot with the kernel image initrd, and NFS mount the rootimg that was generated using 'genimage', I am getting the following error on the compute node's console - FATAL error: could not get the entries from litefile table... after going thru the init-scripts, I found out 'xCATCmd' binary is not present in the rootimg. I am currently checking the xcat packages for its availability. If you know the procedure to get it onto the compute node, please let me know the same. Appreciate your support. Thanking you, Sunil On 6/8/11 9:02 AM, Josh Thompson wrote: Sunil, I don't recall seeing any documentation on those parts. I had to poke around looking at parts of xCAT to see how it worked. It's been a few years since I did that; so, I don't remember much about the process. My recommendation would be to start looking at things in the rootimg.gz image. Looking at it now, I see that /opt/xcat/xcatdsklspost gets run when rootimg.gz boots. It looks like it downloads all of the postscripts from the management node and then run getpostscript.awk which issues a command to xcatd to get the primary postscript for that machine. I've forgotten how xcatd then builds the primary postscript. I do remember that in the partimageng.pm module, I had it add the partimageng postscript. So, you'll really have to start digging through how the xcat postscript system works. Josh On Tuesday June 07, 2011, Sunil Venkatesh wrote: Josh, Is there any place I could find some details on ... /Once the compute node is booted with the stateless image, it uses NFS to mount some things from the management node, and then runs some xcat postscripts,/ I have the stateless images ready with partimage compiled for PPC. For the compute node (power 7) to boot using the stateless images, i need to configure the yaboot instead of pxeboot (which is specific to x86). I wanted to know where in the startup files the execution of partimage and NFS mount is configured. Is it configured by the genimage command itself? Considering the way in which the nodes are configured in the network, it would not be a good idea to let xcat take care of configuring the details like DHCPD for netboot. So, I need to make changes to the configuration files manually, which is why this query came up. Thanks in advance. Regards, Sunil On 6/1/11 1:39 PM, Josh Thompson wrote: Sunil, The stateless image I refer to is what is actually booted on the compute node containing the image to be captured. It's called stateless because it is loaded completely in RAM and does not maintain any state when a reboot occurs. The partimage binary is part of this stateless image and actually runs on the compute node. It does not run on the management node. The management node does not have block level access to the disk on the compute node to be able to capture the image from the disk. I'll try to describe the process a little better. The management node issues a reboot command to the compute node. The compute node uses PXE to load and boot a kernel (vmlinuz), initial RAM disk (initrd.img), and a root filesystem (rootimg.gz) from the management node. All three of these together make up the stateless image. Once the compute node is booted with the stateless image, it uses NFS to mount some things from the management node, and then runs some xcat postscripts, one of which is the
Re: [VCL 2.2.1] [Power7] Problem with image reservation
Sunil, I don't recall seeing any documentation on those parts. I had to poke around looking at parts of xCAT to see how it worked. It's been a few years since I did that; so, I don't remember much about the process. My recommendation would be to start looking at things in the rootimg.gz image. Looking at it now, I see that /opt/xcat/xcatdsklspost gets run when rootimg.gz boots. It looks like it downloads all of the postscripts from the management node and then run getpostscript.awk which issues a command to xcatd to get the primary postscript for that machine. I've forgotten how xcatd then builds the primary postscript. I do remember that in the partimageng.pm module, I had it add the partimageng postscript. So, you'll really have to start digging through how the xcat postscript system works. Josh On Tuesday June 07, 2011, Sunil Venkatesh wrote: Josh, Is there any place I could find some details on ... /Once the compute node is booted with the stateless image, it uses NFS to mount some things from the management node, and then runs some xcat postscripts,/ I have the stateless images ready with partimage compiled for PPC. For the compute node (power 7) to boot using the stateless images, i need to configure the yaboot instead of pxeboot (which is specific to x86). I wanted to know where in the startup files the execution of partimage and NFS mount is configured. Is it configured by the genimage command itself? Considering the way in which the nodes are configured in the network, it would not be a good idea to let xcat take care of configuring the details like DHCPD for netboot. So, I need to make changes to the configuration files manually, which is why this query came up. Thanks in advance. Regards, Sunil On 6/1/11 1:39 PM, Josh Thompson wrote: Sunil, The stateless image I refer to is what is actually booted on the compute node containing the image to be captured. It's called stateless because it is loaded completely in RAM and does not maintain any state when a reboot occurs. The partimage binary is part of this stateless image and actually runs on the compute node. It does not run on the management node. The management node does not have block level access to the disk on the compute node to be able to capture the image from the disk. I'll try to describe the process a little better. The management node issues a reboot command to the compute node. The compute node uses PXE to load and boot a kernel (vmlinuz), initial RAM disk (initrd.img), and a root filesystem (rootimg.gz) from the management node. All three of these together make up the stateless image. Once the compute node is booted with the stateless image, it uses NFS to mount some things from the management node, and then runs some xcat postscripts, one of which is the partimageng postscript. This postscript determines what partitions are on the compute node and, depending on how the postscript is configured, uses partimage or partimageng to capture an image of the compute node disk that is then saved to the management node. When it is finished capturing the image, it notifies xcat on the management node and then reboots. xcat reconfigures itself to tell the compute node to boot off of disk at next boot. When the compute node comes up, it uses PXE to ask the management node how to boot. The management node tells it to boot off of disk. I hope that clarifies how the system works. If any of it is unclear, please ask for further clarification. Josh On Wednesday June 01, 2011, Sunil Venkatesh wrote: Josh, I had one more clarification. partimage binaries run in the management node to capture an (stateless) image from the compute node right? In that case, is there a need for these binaries to go into the rootimg.gz?? My assumption is, partimage runs on the management node (an intel blade in our case) to capture a stateless image from a compute node (a power 7 blade) and stores these images under /install of the management node. Please correct me if I am wrong here. Regards, Sunil On 6/1/11 9:58 AM, Josh Thompson wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Tuesday May 31, 2011, Sunil Venkatesh wrote: Hi, I used the steps that were mentioned under https://cwiki.apache.org/confluence/display/VCL/Adding+support+for+par ti mag e+and+partimage- ng+to+xCAT+2.x+%28unofficial%29 to enable partimage support for xcat. I wasn't sure if I need to change references to x86 x86_64 (as directories) to reflect the ppc architecture, as the web page says The architecture for the node must always be set to x86 for this... I have with me the vmlinuz (kernel image) and initrd for the capture process. The 2 nodeset commands By this, do you mean you have vmlinuz and initrd for your power blades, not the ones linked to off of the page you
Re: [VCL 2.2.1] [Power7] Problem with image reservation
Josh, Is there any place I could find some details on ... /Once the compute node is booted with the stateless image, it uses NFS to mount some things from the management node, and then runs some xcat postscripts,/ I have the stateless images ready with partimage compiled for PPC. For the compute node (power 7) to boot using the stateless images, i need to configure the yaboot instead of pxeboot (which is specific to x86). I wanted to know where in the startup files the execution of partimage and NFS mount is configured. Is it configured by the genimage command itself? Considering the way in which the nodes are configured in the network, it would not be a good idea to let xcat take care of configuring the details like DHCPD for netboot. So, I need to make changes to the configuration files manually, which is why this query came up. Thanks in advance. Regards, Sunil On 6/1/11 1:39 PM, Josh Thompson wrote: Sunil, The stateless image I refer to is what is actually booted on the compute node containing the image to be captured. It's called stateless because it is loaded completely in RAM and does not maintain any state when a reboot occurs. The partimage binary is part of this stateless image and actually runs on the compute node. It does not run on the management node. The management node does not have block level access to the disk on the compute node to be able to capture the image from the disk. I'll try to describe the process a little better. The management node issues a reboot command to the compute node. The compute node uses PXE to load and boot a kernel (vmlinuz), initial RAM disk (initrd.img), and a root filesystem (rootimg.gz) from the management node. All three of these together make up the stateless image. Once the compute node is booted with the stateless image, it uses NFS to mount some things from the management node, and then runs some xcat postscripts, one of which is the partimageng postscript. This postscript determines what partitions are on the compute node and, depending on how the postscript is configured, uses partimage or partimageng to capture an image of the compute node disk that is then saved to the management node. When it is finished capturing the image, it notifies xcat on the management node and then reboots. xcat reconfigures itself to tell the compute node to boot off of disk at next boot. When the compute node comes up, it uses PXE to ask the management node how to boot. The management node tells it to boot off of disk. I hope that clarifies how the system works. If any of it is unclear, please ask for further clarification. Josh On Wednesday June 01, 2011, Sunil Venkatesh wrote: Josh, I had one more clarification. partimage binaries run in the management node to capture an (stateless) image from the compute node right? In that case, is there a need for these binaries to go into the rootimg.gz?? My assumption is, partimage runs on the management node (an intel blade in our case) to capture a stateless image from a compute node (a power 7 blade) and stores these images under /install of the management node. Please correct me if I am wrong here. Regards, Sunil On 6/1/11 9:58 AM, Josh Thompson wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Tuesday May 31, 2011, Sunil Venkatesh wrote: Hi, I used the steps that were mentioned under https://cwiki.apache.org/confluence/display/VCL/Adding+support+for+parti mag e+and+partimage- ng+to+xCAT+2.x+%28unofficial%29 to enable partimage support for xcat. I wasn't sure if I need to change references to x86 x86_64 (as directories) to reflect the ppc architecture, as the web page says The architecture for the node must always be set to x86 for this... I have with me the vmlinuz (kernel image) and initrd for the capture process. The 2 nodeset commands By this, do you mean you have vmlinuz and initrd for your power blades, not the ones linked to off of the page you listed above? If you do, that's a good start. However, you'll also need rootimg.gz. rootimg.gz is the root filesystem for the stateless image. It also contains the partimage and partimageng binaries. Assuming partimage or partimageng can actually capture partitions from power systems, you'll need to compile at least one of them to run on power. For the rootimg.gz image I provided, I compiled them statically so that I didn't have to worry about including any library dependencies in rootimg.gz. It would be a good idea to research how to use xcat's genimage command to generate stateless images to learn how to do this. If there's any part of the above that you don't fully understand, please ask me to clarify it. Until you have a stateless image that you can deploy to your power blades, there's no point in trying to debug any VCL specific items. Josh - -- - --- Josh Thompson VCL Developer North Carolina State University my GPG/PGP key can be found at pgp.mit.edu -BEGIN PGP
Re: [VCL 2.2.1] [Power7] Problem with image reservation
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Tuesday May 31, 2011, Sunil Venkatesh wrote: Hi, I used the steps that were mentioned under https://cwiki.apache.org/confluence/display/VCL/Adding+support+for+partimag e+and+partimage- ng+to+xCAT+2.x+%28unofficial%29 to enable partimage support for xcat. I wasn't sure if I need to change references to x86 x86_64 (as directories) to reflect the ppc architecture, as the web page says The architecture for the node must always be set to x86 for this... I have with me the vmlinuz (kernel image) and initrd for the capture process. The 2 nodeset commands By this, do you mean you have vmlinuz and initrd for your power blades, not the ones linked to off of the page you listed above? If you do, that's a good start. However, you'll also need rootimg.gz. rootimg.gz is the root filesystem for the stateless image. It also contains the partimage and partimageng binaries. Assuming partimage or partimageng can actually capture partitions from power systems, you'll need to compile at least one of them to run on power. For the rootimg.gz image I provided, I compiled them statically so that I didn't have to worry about including any library dependencies in rootimg.gz. It would be a good idea to research how to use xcat's genimage command to generate stateless images to learn how to do this. If there's any part of the above that you don't fully understand, please ask me to clarify it. Until you have a stateless image that you can deploy to your power blades, there's no point in trying to debug any VCL specific items. Josh - -- - --- Josh Thompson VCL Developer North Carolina State University my GPG/PGP key can be found at pgp.mit.edu -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.17 (GNU/Linux) iEYEARECAAYFAk3mRYsACgkQV/LQcNdtPQNnVgCbB9ZFJn0+C45RC/g75RqGZY/j PZYAniP2Eam7nxgiDWUnp5sKPYPO4OMa =exBV -END PGP SIGNATURE-
Re: [VCL 2.2.1] [Power7] Problem with image reservation
Josh, Thank you for that detailed clarification. Appreciate your support. Regards, Sunil On 6/1/11 1:39 PM, Josh Thompson wrote: Sunil, The stateless image I refer to is what is actually booted on the compute node containing the image to be captured. It's called stateless because it is loaded completely in RAM and does not maintain any state when a reboot occurs. The partimage binary is part of this stateless image and actually runs on the compute node. It does not run on the management node. The management node does not have block level access to the disk on the compute node to be able to capture the image from the disk. I'll try to describe the process a little better. The management node issues a reboot command to the compute node. The compute node uses PXE to load and boot a kernel (vmlinuz), initial RAM disk (initrd.img), and a root filesystem (rootimg.gz) from the management node. All three of these together make up the stateless image. Once the compute node is booted with the stateless image, it uses NFS to mount some things from the management node, and then runs some xcat postscripts, one of which is the partimageng postscript. This postscript determines what partitions are on the compute node and, depending on how the postscript is configured, uses partimage or partimageng to capture an image of the compute node disk that is then saved to the management node. When it is finished capturing the image, it notifies xcat on the management node and then reboots. xcat reconfigures itself to tell the compute node to boot off of disk at next boot. When the compute node comes up, it uses PXE to ask the management node how to boot. The management node tells it to boot off of disk. I hope that clarifies how the system works. If any of it is unclear, please ask for further clarification. Josh On Wednesday June 01, 2011, Sunil Venkatesh wrote: Josh, I had one more clarification. partimage binaries run in the management node to capture an (stateless) image from the compute node right? In that case, is there a need for these binaries to go into the rootimg.gz?? My assumption is, partimage runs on the management node (an intel blade in our case) to capture a stateless image from a compute node (a power 7 blade) and stores these images under /install of the management node. Please correct me if I am wrong here. Regards, Sunil On 6/1/11 9:58 AM, Josh Thompson wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Tuesday May 31, 2011, Sunil Venkatesh wrote: Hi, I used the steps that were mentioned under https://cwiki.apache.org/confluence/display/VCL/Adding+support+for+parti mag e+and+partimage- ng+to+xCAT+2.x+%28unofficial%29 to enable partimage support for xcat. I wasn't sure if I need to change references to x86 x86_64 (as directories) to reflect the ppc architecture, as the web page says The architecture for the node must always be set to x86 for this... I have with me the vmlinuz (kernel image) and initrd for the capture process. The 2 nodeset commands By this, do you mean you have vmlinuz and initrd for your power blades, not the ones linked to off of the page you listed above? If you do, that's a good start. However, you'll also need rootimg.gz. rootimg.gz is the root filesystem for the stateless image. It also contains the partimage and partimageng binaries. Assuming partimage or partimageng can actually capture partitions from power systems, you'll need to compile at least one of them to run on power. For the rootimg.gz image I provided, I compiled them statically so that I didn't have to worry about including any library dependencies in rootimg.gz. It would be a good idea to research how to use xcat's genimage command to generate stateless images to learn how to do this. If there's any part of the above that you don't fully understand, please ask me to clarify it. Until you have a stateless image that you can deploy to your power blades, there's no point in trying to debug any VCL specific items. Josh - -- - --- Josh Thompson VCL Developer North Carolina State University my GPG/PGP key can be found at pgp.mit.edu -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.17 (GNU/Linux) iEYEARECAAYFAk3mRYsACgkQV/LQcNdtPQNnVgCbB9ZFJn0+C45RC/g75RqGZY/j PZYAniP2Eam7nxgiDWUnp5sKPYPO4OMa =exBV -END PGP SIGNATURE-
Re: [VCL 2.2.1] [Power7] Problem with image reservation
Hi, I used the steps that were mentioned under https://cwiki.apache.org/confluence/display/VCL/Adding+support+for+partimage+and+partimage- ng+to+xCAT+2.x+%28unofficial%29 to enable partimage support for xcat. I wasn't sure if I need to change references to x86 x86_64 (as directories) to reflect the ppc architecture, as the web page says The architecture for the node must always be set to x86 for this... I have with me the vmlinuz (kernel image) and initrd for the capture process. The 2 nodeset commands nodeset node image nodeset node install fail with the following error msg: Error: Invalid nodes and/or groups in noderange: power01. Also, when I start the capture process using vcld --setup, again, there is no option to choose a ppc architecture (I am not sure if I should be modifying the setup scripts to support ppc architecture). I continue with the setup process by choosing the architecture as x86_64 (instead of ppc for our Power 7 blade). Operating System is RHEL 5. The scripts attempts to power down the blade, waits for 120 seconds before it quits with a failure. The log file is shown at the bottom of the mail. The script fails saying the blade didn't turn off, but, I can see in my console that the blade is off and it does not respond to connection requests. There is a deviation from the configuration that were mentioned: the blade we are trying to use as a compute node has only 1 ethernet port enabled, hence there is no separate private public networks. Will that matter? Any support with this would be really helpful. Thanks in advance. Regards, Sunil Venkatesh *** RECENT LOG ENTRIES FOR THIS PROCESS: |11662|23:23|image| (-2) Module.pm, code_loop_timeout (line: 761) |11662|23:23|image| (-3) Provisioning.pm, wait_for_power_off (line: 324) |11662|23:23|image| (-4) Linux.pm, pre_capture (line: 183) |11662|23:23|image| (-5) xCAT2.pm, capture (line: 792) 2011-05-31 12:56:37|11662|23:23|image|Module.pm:code_loop_timeout(755)|attempt 36: code returned false, seconds elapsed/remaining: 114/6, sleeping for 3 seconds 2011-05-31 12:56:40|11662|23:23|image|Module.pm:code_loop_timeout(759)|attempt 37: waiting for power01 to power off 2011-05-31 12:56:40|11662|23:23|image|xCAT.pm:_rpower(1944)|attempting to execute rpower for computer: power01, mode: stat 2011-05-31 12:56:40|11662|23:23|image|utils.pm:run_command(9010)|executed command: /opt/xcat/bin/rpower power01 stat, pid: 11922, exit status: 1, output: |11662|23:23|image| Error: Invalid nodes and/or groups in noderange: power01 |11662|23:23|image| WARNING |11662|23:23|image| 2011-05-31 12:56:40|11662|23:23|image|xCAT.pm:_rpower(1997)|unexpected output returned from rpower: Error: Invalid nodes and/or groups in noderange: power01 |11662|23:23|image| ( 0) xCAT.pm, _rpower (line: 1997) |11662|23:23|image| (-1) xCAT.pm, power_status (line: 1675) |11662|23:23|image| (-2) Provisioning.pm, __ANON__ (line: 324) |11662|23:23|image| (-3) Module.pm, code_loop_timeout (line: 761) |11662|23:23|image| (-4) Provisioning.pm, wait_for_power_off (line: 324) |11662|23:23|image| (-5) Linux.pm, pre_capture (line: 183) 2011-05-31 12:56:40|11662|23:23|image|xCAT.pm:power_status(1676)|retrieved power status of power01: 0 |11662|23:23|image| WARNING |11662|23:23|image| 2011-05-31 12:56:40|11662|23:23|image|xCAT.pm:power_status(1679)|failed to determine power status, rpower subroutine returned 0 |11662|23:23|image| ( 0) xCAT.pm, power_status (line: 1679) |11662|23:23|image| (-1) Provisioning.pm, __ANON__ (line: 324) |11662|23:23|image| (-2) Module.pm, code_loop_timeout (line: 761) |11662|23:23|image| (-3) Provisioning.pm, wait_for_power_off (line: 324) |11662|23:23|image| (-4) Linux.pm, pre_capture (line: 183) |11662|23:23|image| (-5) xCAT2.pm, capture (line: 792) |11662|23:23|image| WARNING |11662|23:23|image| 2011-05-31 12:56:40|11662|23:23|image|vcld:warning_handler(610)|Use of uninitialized value in pattern match (m//) at /usr/local/vcl/bin/../lib/VCL/Module/Provisioning.pm line 324. |11662|23:23|image| ( 0) vcld, warning_handler (line: 610) |11662|23:23|image| (-1) Provisioning.pm, __ANON__ (line: 324) |11662|23:23|image| (-2) Module.pm, code_loop_timeout (line: 761) |11662|23:23|image| (-3) Provisioning.pm, wait_for_power_off (line: 324) |11662|23:23|image| (-4) Linux.pm, pre_capture (line: 183) |11662|23:23|image| (-5) xCAT2.pm, capture (line: 792) 2011-05-31 12:56:40|11662|23:23|image|Module.pm:code_loop_timeout(755)|attempt 37: code returned false, seconds elapsed/remaining: 117/3, sleeping for 3 seconds 2011-05-31 12:56:43|11662|23:23|image|Module.pm:code_loop_timeout(759)|attempt 38: waiting for power01 to power off 2011-05-31 12:56:43|11662|23:23|image|xCAT.pm:_rpower(1944)|attempting to execute rpower for computer: power01, mode: stat 2011-05-31 12:56:44|11662|23:23|image|utils.pm:run_command(9010)|executed command: /opt/xcat/bin/rpower power01 stat, pid: 11926, exit
[VCL 2.2.1] [Power7] Problem with image reservation
Hi, We are currently in the process of configuring VCL 2.2.1 to work on a Power 7 blade. Our current setup is: 1. A web-server that hosts the Database and the Web Code. The same server acts as the Management node. xCAT is configured as the provisioning module on this node. 2. Power7 is our compute node. 3. I used the command vcld --setup command to create/capture base image of RHEL 5 that is running on the Power7 blade (by specifying the IP address of Power7 blade when prompted for an address). The creation process failed as Xianqing Yu had mentioned to us earlier. Although, before it failed it created appropriate entries in the tables image, imagerevision and resource. I was able to Undelete the image from the web page and see it under New Reservations. I am facing similar problems that Mike Waldron had faced with the reservation. Even after making memory adjustment, I wasn't able to make a reservation. The time table shows all green (available), however, when I choose any entry from the list, it takes me directly to New Reservation page without any status/feedback. And, I don't see any reservations created when I check under Current Reservations. I am just assuming the groupings of Images and Computers are correct, is there anyway I could verify the same. Also, if there is any reference to how the grouping need to be done, please let me know of the same. Please do correct me if there is anything wrong with the system setup. Regards, Sunil Venkatesh Research Assistant, MC2 Lab, UMBC.
Re: [VCL 2.2.1] [Power7] Problem with image reservation
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Sunil, Let's back up a little bit. The first thing to look at is why the image failed. Unless you created your own stateless image for capturing and provisioning images, then xCAT will be unable to capture an image from a Power blade. Did you use the steps here: https://cwiki.apache.org/confluence/display/VCL/Adding+support+for+partimage+and+partimage- ng+to+xCAT+2.x+%28unofficial%29 for modifying xCAT to be able to capture/deploy images? If so, the stateless images linked to off of that page are for x86 hardware. You will need to create your own stateless or statelite images for Power blades. Josh On Thursday May 19, 2011, Sunil Venkatesh wrote: Hi, We are currently in the process of configuring VCL 2.2.1 to work on a Power 7 blade. Our current setup is: 1. A web-server that hosts the Database and the Web Code. The same server acts as the Management node. xCAT is configured as the provisioning module on this node. 2. Power7 is our compute node. 3. I used the command vcld --setup command to create/capture base image of RHEL 5 that is running on the Power7 blade (by specifying the IP address of Power7 blade when prompted for an address). The creation process failed as Xianqing Yu had mentioned to us earlier. Although, before it failed it created appropriate entries in the tables image, imagerevision and resource. I was able to Undelete the image from the web page and see it under New Reservations. I am facing similar problems that Mike Waldron had faced with the reservation. Even after making memory adjustment, I wasn't able to make a reservation. The time table shows all green (available), however, when I choose any entry from the list, it takes me directly to New Reservation page without any status/feedback. And, I don't see any reservations created when I check under Current Reservations. I am just assuming the groupings of Images and Computers are correct, is there anyway I could verify the same. Also, if there is any reference to how the grouping need to be done, please let me know of the same. Please do correct me if there is anything wrong with the system setup. Regards, Sunil Venkatesh Research Assistant, MC2 Lab, UMBC. - -- - --- Josh Thompson VCL Developer North Carolina State University my GPG/PGP key can be found at pgp.mit.edu -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.17 (GNU/Linux) iEYEARECAAYFAk3VdBQACgkQV/LQcNdtPQP5VQCfVoX4ykJSCpMHHJTocpwTHsVs teEAn2NCYnBXDq/gzjcwj2FNn9kdJsPC =COKC -END PGP SIGNATURE-
