from:"Bernard Li"

Hi Franck:

On Thu, Apr 14, 2011 at 9:40 AM, E-Blokos in...@e-blokos.com wrote:

 arrgh, didn't remember and didn't note anything.
 sorry for my stupid question but how to make a screenshot/dump at this
 when the computer is in this state ?
 btw, I planned to install another node soon, will see if it does the same
 so I will read carefully what's happen

You can try setting up si_monitortk to see if it catches the messages.
 If not and you don't have console access at all, the best bet is
probably to take a digital picture of your screen with the text
focused such that you can see what's going on.

I will be posting the most recent snapshot with ext4 support built-in
shortly, perhaps you can wait to test that instead.

Cheers,

Bernard

--
Benefiting from Server Virtualization: Beyond Initial Workload 
Consolidation -- Increasing the use of server virtualization is a top
priority.Virtualization can reduce costs, simplify management, and improve 
application availability and disaster protection. Learn more about boosting 
the value of server virtualization. http://p.sf.net/sfu/vmware-sfdev2dev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

[sisuite-users] Latest snapshot for systemimager 4.2.0 available with ext4 support

Hi all:

The latest snapshot for SystemImager 4.2.0 is now available:

http://download.systemimager.org/~bli/systemimager/

Changes since the last snapshot:

- added bc, rsync = 2.4.6 to RPM spec file BuildRequires
- fix 'Use of uninitialized value' warning and disallow passing empty
string to --update-script for si_getimage
- allow parted version numbers a.b in addition to a.b.c
- added support of openSUSE 11.3 kernels in _get_kernel_release() for UYOK
- added support for devtmpfs devstyle (openSUSE 11.3+)
- added support for ext4 filesystem

Thanks for the contributions of Thomas Zeiser and Franck.

If you could please test the latest snapshot and provide feedback,
that would be greatly appreciated.

Cheers,

Bernard

--
Benefiting from Server Virtualization: Beyond Initial Workload 
Consolidation -- Increasing the use of server virtualization is a top
priority.Virtualization can reduce costs, simplify management, and improve 
application availability and disaster protection. Learn more about boosting 
the value of server virtualization. http://p.sf.net/sfu/vmware-sfdev2dev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] si_clusterconfig

Hi Franck:

On Fri, Apr 8, 2011 at 9:59 PM, E-Blokos in...@e-blokos.com wrote:

 when I use ip-rangenode1-node2/ip-range
 it gives Perl error

 Caching XML configuration... Use of uninitialized value
 $etc_hosts_lines_by_number{13} in concatenation (.) or string at
 /usr//lib/systemimager/perl/SystemImager/HostRange.pm line 352

 I should use a comma with another node to remove the error like
 ip-rangenode0,node1-node2/ip-range

I think you're doing this incorrectly.  If you want to specify host
range, you should be using the node/node syntax.  For
ip-range/ip-range, you should provide it with ip addresses, not
hostnames.

Granted, there should probably be better sanity/error checking for the script.

Cheers,

Bernard

--
Benefiting from Server Virtualization: Beyond Initial Workload 
Consolidation -- Increasing the use of server virtualization is a top
priority.Virtualization can reduce costs, simplify management, and improve 
application availability and disaster protection. Learn more about boosting 
the value of server virtualization. http://p.sf.net/sfu/vmware-sfdev2dev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Latest snapshot for systemimager 4.2.0 availablewith ext4 support

On Thu, Apr 14, 2011 at 11:48 AM, E-Blokos in...@e-blokos.com wrote:

 I can see only 4.1.99.svn4555 package on the link above

Yes, that's the 4.2.0 snapshot -- am I missing something?

Regards,

Bernard

--
Benefiting from Server Virtualization: Beyond Initial Workload 
Consolidation -- Increasing the use of server virtualization is a top
priority.Virtualization can reduce costs, simplify management, and improve 
application availability and disaster protection. Learn more about boosting 
the value of server virtualization. http://p.sf.net/sfu/vmware-sfdev2dev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Help test latest snapshot of SystemImager

2011-04-13 Thread Bernard Li

Hi Franck:

On Tue, Apr 12, 2011 at 9:24 AM, E-Blokos in...@e-blokos.com wrote:

 - after the image was installed before reboot a kernel panic occured

Can you please provide more information regarding this?  What module
caused the panic?  A screenshot/dump of the screen would be helpful.

 - even if in systemimager.conf NET_BOOT_DEFAULT was set in local at reboot
 it retried to reinstall the image
 - the UUIDs of all HD failed (seems the module uuid didn't load) so I
 modified /etc/mdadm.conf to match the real /dev/sdX in the ARRAY lines
 - I had to change manually the network-scripts
 - I needed to boot first in rescue mode to rebuild all the Raid1 arrays

I suspect you are getting the above weird issues due to the kernel
panic and SystemImager not being able to complete successfully.

Are other users experiencing similar issues?  If you are able to test
this release, please report back pass/fail.  Any feedback would help
with making this release available officially sooner.

Thanks!

Bernard

--
Forrester Wave Report - Recovery time is now measured in hours and minutes
not days. Key insights are discussed in the 2010 Forrester Wave Report as
part of an in-depth evaluation of disaster recovery service providers.
Forrester found the best-in-class provider in terms of services and vision.
Read this report now!  http://p.sf.net/sfu/ibm-webcastpromo
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Help test latest snapshot of SystemImager

2011-04-12 Thread Bernard Li

Hi Thomas:

On Tue, Apr 12, 2011 at 1:33 PM, Thomas Zeiser
thomas.zei...@rrze.uni-erlangen.de wrote:

 Just one remark: your implementation for @528/530 makes the
 default behavior for update_script equal to updating the script,
 i.e.  if no --update_script is given, it is equivalent to
 --update_script=yes as the exit is not executed. Is that the
 expected behavior? I do not know how often people customize their
 image.script by hand after it was initially created and who do not
 want to have it overwritten accidentially.

Right...  in that case I guess your original patch is the correct one
-- will check that in.

Thanks,

Bernard

--
Forrester Wave Report - Recovery time is now measured in hours and minutes
not days. Key insights are discussed in the 2010 Forrester Wave Report as
part of an in-depth evaluation of disaster recovery service providers.
Forrester found the best-in-class provider in terms of services and vision.
Read this report now!  http://p.sf.net/sfu/ibm-webcastpromo
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Help test latest snapshot of SystemImager

2011-04-12 Thread Bernard Li

Hi Franck:

On Tue, Apr 12, 2011 at 9:24 AM, E-Blokos in...@e-blokos.com wrote:

 my experience with your last version (4.1.99 r4537) has been a success with
 a UYOK because
 apparently the standard kernel doesn't include module for adaptec sata card.
 so after netbooted and automatically installed a new node I noticed :
 - after the image was installed before reboot a kernel panic occured
 - even if in systemimager.conf NET_BOOT_DEFAULT was set in local at reboot
 it retried to reinstall the image
 - the UUIDs of all HD failed (seems the module uuid didn't load) so I
 modified /etc/mdadm.conf to match the real /dev/sdX in the ARRAY lines
 - I had to change manually the network-scripts
 - I needed to boot first in rescue mode to rebuild all the Raid1 arrays

So are you saying that you encounter the above problems with the
*standard* kernel but with UYOK, everything was smooth sailing with no
issues?

Thomas, did you test with standard or UYOK kernel/initrd?

Thanks,

Bernard

--
Forrester Wave Report - Recovery time is now measured in hours and minutes
not days. Key insights are discussed in the 2010 Forrester Wave Report as
part of an in-depth evaluation of disaster recovery service providers.
Forrester found the best-in-class provider in terms of services and vision.
Read this report now!  http://p.sf.net/sfu/ibm-webcastpromo
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Help test latest snapshot of SystemImager

Hi Andy:

On Mon, Apr 11, 2011 at 6:40 AM, Andy Rabagliati an...@wizzy.com wrote:

 I am having difficulty resolving this site.

 Can anyone give me an IP address ?

 The nameservers for systemimager.org - listed as
 ns1.thefinleys.com and ns2.thefinleys.com also do not resolve.

 whois data for  systemimager.org and thefinleys.com does not
 list IP addresses for their nameservers..

Where are you based?  Perhaps there are some issues with your ISP's DNS servers?

Cheers,

Bernard

--
Xperia(TM) PLAY
It's a major breakthrough. An authentic gaming
smartphone on the nation's most reliable network.
And it wants your games.
http://p.sf.net/sfu/verizon-sfdev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Help test latest snapshot of SystemImager

Hi Thomas:

Please test the following *untested* patch:

Index: lib/SystemImager/Server.pm
===
--- lib/SystemImager/Server.pm  (revision 4539)
+++ lib/SystemImager/Server.pm  (working copy)
@@ -1760,7 +1760,8 @@
 $xml_config = XMLin($file, keyattr = { boel = +devstyle} );
 if( defined($xml_config-{boel}-{devstyle})
  (($xml_config-{boel}-{devstyle} eq udev )
- or ($xml_config-{boel}-{devstyle} eq devfs) )
+ or ($xml_config-{boel}-{devstyle} eq devfs)
+ or ($xml_config-{boel}-{devstyle} eq devtmpfs) )
   ) {
 $fs_by_mp{'/dev'} = /dev;
 }
@@ -2316,6 +2317,7 @@
 #   Clients should have one of the following entries in their
 #   autoinstallscript.conf file:
 #
+#   boel devstyle=devtmpfs/
 #   boel devstyle=udev/
 #   boel devstyle=devfs/
 #   boel devstyle=static/
@@ -2332,7 +2334,8 @@

 if( defined($xml_config-{boel}-{devstyle})
  (($xml_config-{boel}-{devstyle} eq devfs)
- or ($xml_config-{boel}-{devstyle} eq udev ) )
+ or ($xml_config-{boel}-{devstyle} eq udev )
+ or ($xml_config-{boel}-{devstyle} eq devtmpfs ) )
   ) {

 my $cmd = q(mount /dev /a/dev -o bind || shellout);

This, along with your patch against lib/SystemImager/Common.pm should
add support for devtmpfs.  Please test it out and let me know if it
works or not.  You'll need to re-generate autoinstallscript.conf and
the master script respectively.

Cheers,

Bernard

On Mon, Apr 11, 2011 at 11:59 AM, Thomas Zeiser
thomas.zei...@rrze.uni-erlangen.de wrote:
 Hi Bernard,

 On Fri, Apr 08, 2011 at 12:35:58PM -0700, Bernard Li wrote:
 Hi Thomas:

 I'm a bit hesitant to make such a major change to add support for
 devtmpfs -- does imaging work correctly if you enable UYOK?

 I fully understand that you hesitate. However, if the client OS
 uses devtmpfs neither UYOK nor the standard kernel will work
 without manually editing the image.master script because
 /etc/systemimager/autoinstallscript.conf will have beol
 devstyle=static/ (which is completely wrong). As a consequnce,
 imager.master won't get a mount /dev /a/dev -o bin  line. Thus,
 /a/dev is not populated and systemconfigurator won't find the
 harddisk to initialize grub, etc.

 Thus, I highly would appreciate any fix that recognizes devtmpfs
 and maks the image.master script mount /a/dev. Or at least a big
 warning in the documentation that this line has be added manually.
 A more general alternative might be adding testing code to
 image.master to see if /a/dev looks reasonable populated OR mount
 /a/dev if not.

 Best,

 thomas

 Cheers,

 Bernard

 On Fri, Apr 8, 2011 at 11:48 AM, Thomas Zeiser
 thomas.zei...@rrze.uni-erlangen.de wrote:
  Hi Bernard,
 
  here is one more:
 
  OpenSuSE 11.3 does no longer uses devfs but devtmpfs ...
  An example of /proc/mounts
 
  rootfs / rootfs rw 0 0
  devtmpfs /dev devtmpfs
  rw,relatime,size=186100k,nr_inodes=46525,mode=755 0 0
  tmpfs /dev/shm tmpfs rw,relatime 0 0
  devpts /dev/pts devpts rw,relatime,gid=5,mode=620,ptmxmode=000 0 0
  /dev/sda2 / ext4 rw,relatime,user_xattr,acl,barrier=1,data=ordered 0 0
  proc /proc proc rw,relatime 0 0
  sysfs /sys sysfs rw,relatime 0 0
  debugfs /sys/kernel/debug debugfs rw,relatime 0 0
 
  Common.pm won't detect detect that and set boel devstyle to
  standard instead of udev (and thus not mount /a/dev).
 
 
  The attached patch extends the regular pattern. But that does not
  really solve the problem. If no-uyok is used, it's not important
  what the running system uses but what the standard kernel will
  require during install.  The new standard kernel will always
  require boel devsytle=udev (or similar) to have /a/dev mounted
  before systemconfigurator will be called from within the
  master.script.
 
 
  Best,
 
  thomas


--
Forrester Wave Report - Recovery time is now measured in hours and minutes
not days. Key insights are discussed in the 2010 Forrester Wave Report as
part of an in-depth evaluation of disaster recovery service providers.
Forrester found the best-in-class provider in terms of services and vision.
Read this report now!  http://p.sf.net/sfu/ibm-webcastpromo
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Help test latest snapshot of SystemImager

You probably need this patch too:

Index: lib/SystemImager/Common.pm
===
--- lib/SystemImager/Common.pm  (revision 4539)
+++ lib/SystemImager/Common.pm  (working copy)
@@ -1669,6 +1669,8 @@
 return 'udev';
 } elsif (m/\bdevfs\b/) {
 return 'devfs';
+if ((m/\bdevtmpfs\b/) || (m/\/dev\sdevtmpfs\s/)) {
+return 'devtmpfs';
 }
 }
 close(FILE);

Cheers,

Bernard

On Mon, Apr 11, 2011 at 12:16 PM, Bernard Li bern...@vanhpc.org wrote:
 Hi Thomas:

 Please test the following *untested* patch:

 Index: lib/SystemImager/Server.pm
 ===
 --- lib/SystemImager/Server.pm  (revision 4539)
 +++ lib/SystemImager/Server.pm  (working copy)
 @@ -1760,7 +1760,8 @@
     $xml_config = XMLin($file, keyattr = { boel = +devstyle} );
     if( defined($xml_config-{boel}-{devstyle})
          (    ($xml_config-{boel}-{devstyle} eq udev )
 -             or ($xml_config-{boel}-{devstyle} eq devfs) )
 +             or ($xml_config-{boel}-{devstyle} eq devfs)
 +             or ($xml_config-{boel}-{devstyle} eq devtmpfs) )
       ) {
         $fs_by_mp{'/dev'} = /dev;
     }
 @@ -2316,6 +2317,7 @@
  #   Clients should have one of the following entries in their
  #   autoinstallscript.conf file:
  #
 +#       boel devstyle=devtmpfs/
  #       boel devstyle=udev/
  #       boel devstyle=devfs/
  #       boel devstyle=static/
 @@ -2332,7 +2334,8 @@

     if( defined($xml_config-{boel}-{devstyle})
          (    ($xml_config-{boel}-{devstyle} eq devfs)
 -             or ($xml_config-{boel}-{devstyle} eq udev ) )
 +             or ($xml_config-{boel}-{devstyle} eq udev )
 +             or ($xml_config-{boel}-{devstyle} eq devtmpfs ) )
       ) {

         my $cmd = q(mount /dev /a/dev -o bind || shellout);

 This, along with your patch against lib/SystemImager/Common.pm should
 add support for devtmpfs.  Please test it out and let me know if it
 works or not.  You'll need to re-generate autoinstallscript.conf and
 the master script respectively.

 Cheers,

 Bernard

 On Mon, Apr 11, 2011 at 11:59 AM, Thomas Zeiser
 thomas.zei...@rrze.uni-erlangen.de wrote:
 Hi Bernard,

 On Fri, Apr 08, 2011 at 12:35:58PM -0700, Bernard Li wrote:
 Hi Thomas:

 I'm a bit hesitant to make such a major change to add support for
 devtmpfs -- does imaging work correctly if you enable UYOK?

 I fully understand that you hesitate. However, if the client OS
 uses devtmpfs neither UYOK nor the standard kernel will work
 without manually editing the image.master script because
 /etc/systemimager/autoinstallscript.conf will have beol
 devstyle=static/ (which is completely wrong). As a consequnce,
 imager.master won't get a mount /dev /a/dev -o bin  line. Thus,
 /a/dev is not populated and systemconfigurator won't find the
 harddisk to initialize grub, etc.

 Thus, I highly would appreciate any fix that recognizes devtmpfs
 and maks the image.master script mount /a/dev. Or at least a big
 warning in the documentation that this line has be added manually.
 A more general alternative might be adding testing code to
 image.master to see if /a/dev looks reasonable populated OR mount
 /a/dev if not.

 Best,

 thomas

 Cheers,

 Bernard

 On Fri, Apr 8, 2011 at 11:48 AM, Thomas Zeiser
 thomas.zei...@rrze.uni-erlangen.de wrote:
  Hi Bernard,
 
  here is one more:
 
  OpenSuSE 11.3 does no longer uses devfs but devtmpfs ...
  An example of /proc/mounts
 
  rootfs / rootfs rw 0 0
  devtmpfs /dev devtmpfs
  rw,relatime,size=186100k,nr_inodes=46525,mode=755 0 0
  tmpfs /dev/shm tmpfs rw,relatime 0 0
  devpts /dev/pts devpts rw,relatime,gid=5,mode=620,ptmxmode=000 0 0
  /dev/sda2 / ext4 rw,relatime,user_xattr,acl,barrier=1,data=ordered 0 0
  proc /proc proc rw,relatime 0 0
  sysfs /sys sysfs rw,relatime 0 0
  debugfs /sys/kernel/debug debugfs rw,relatime 0 0
 
  Common.pm won't detect detect that and set boel devstyle to
  standard instead of udev (and thus not mount /a/dev).
 
 
  The attached patch extends the regular pattern. But that does not
  really solve the problem. If no-uyok is used, it's not important
  what the running system uses but what the standard kernel will
  require during install.  The new standard kernel will always
  require boel devsytle=udev (or similar) to have /a/dev mounted
  before systemconfigurator will be called from within the
  master.script.
 
 
  Best,
 
  thomas



--
Forrester Wave Report - Recovery time is now measured in hours and minutes
not days. Key insights are discussed in the 2010 Forrester Wave Report as
part of an in-depth evaluation of disaster recovery service providers.
Forrester found the best-in-class provider in terms of services and vision.
Read this report now!  http://p.sf.net/sfu/ibm-webcastpromo

Re: [sisuite-users] Help test latest snapshot of SystemImager

Hi Thomas:

Does this work?

Index: sbin/si_getimage
===
--- sbin/si_getimage(revision 4539)
+++ sbin/si_getimage(working copy)
@@ -197,10 +197,12 @@
 }

 # be sure $update_script was passed a proper option
-unless(
-   ($update_script eq yes)
-or ($update_script eq no )
-) { die \n$program_name: --update-script must be yes or no.\n$get_help\n\n; }
+if (defined $update_script) {
+  unless(
+ ($update_script eq yes)
+  or ($update_script eq no )
+  ) { die \n$program_name: --update-script must be yes or
no.\n$get_help\n\n; }
+}

 SystemImager::Server-validate_ip_assignment_option($ip_assignment_method);
 SystemImager::Server-validate_post_install_option($post_install);
@@ -528,7 +530,9 @@
 }

 # if we don't need to update the master script, then exit -- we're done
-unless($update_script eq yes) { exit 0; }
+if (defined $update_script) {
+  unless($update_script eq yes) { exit 0; }
+}

 ### END Overwrite $script_name.master? ###

Thanks,

Bernard

On Fri, Apr 8, 2011 at 7:58 AM, Thomas Zeiser
thomas.zei...@rrze.uni-erlangen.de wrote:
 On Thu, Apr 07, 2011 at 05:11:26PM -0700, Bernard Li wrote:
 This seems to be a better patch, as it also catches if the user
 specified --update-script '' on the CLI:

 Index: sbin/si_getimage
 ===
 --- sbin/si_getimage  (revision 4539)
 +++ sbin/si_getimage  (working copy)
 @@ -197,10 +197,12 @@
  }

  # be sure $update_script was passed a proper option
 -unless(
 -       ($update_script eq yes)
 -    or ($update_script eq no )
 -) { die \n$program_name: --update-script must be yes or 
 no.\n$get_help\n\n; }
 +if (defined $update_script) {
 +  unless(
 +         ($update_script eq yes)
 +      or ($update_script eq no )
 +  ) { die \n$program_name: --update-script must be yes or
 no.\n$get_help\n\n; }
 +}

  SystemImager::Server-validate_ip_assignment_option($ip_assignment_method);
  SystemImager::Server-validate_post_install_option($post_install);

 That's only half of the storry; you also need the following part;

 ==
 @@ -531,7 +531,7 @@
  }

  # if we don't need to update the master script, then exit -- we're
  # done
 -unless($update_script eq yes) { exit 0; }
 +unless((defined($update_script)) and ($update_script eq yes)) { exit 0; }

  ### END Overwrite $script_name.master? ###

 ==


 Best,

 thomas


--
Forrester Wave Report - Recovery time is now measured in hours and minutes
not days. Key insights are discussed in the 2010 Forrester Wave Report as
part of an in-depth evaluation of disaster recovery service providers.
Forrester found the best-in-class provider in terms of services and vision.
Read this report now!  http://p.sf.net/sfu/ibm-webcastpromo
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Help test latest snapshot of SystemImager

2011-04-08 Thread Bernard Li

Hi Thomas:

I'm a bit hesitant to make such a major change to add support for
devtmpfs -- does imaging work correctly if you enable UYOK?

Cheers,

Bernard

On Fri, Apr 8, 2011 at 11:48 AM, Thomas Zeiser
thomas.zei...@rrze.uni-erlangen.de wrote:
 Hi Bernard,

 here is one more:

 OpenSuSE 11.3 does no longer uses devfs but devtmpfs ...
 An example of /proc/mounts

 rootfs / rootfs rw 0 0
 devtmpfs /dev devtmpfs
 rw,relatime,size=186100k,nr_inodes=46525,mode=755 0 0
 tmpfs /dev/shm tmpfs rw,relatime 0 0
 devpts /dev/pts devpts rw,relatime,gid=5,mode=620,ptmxmode=000 0 0
 /dev/sda2 / ext4 rw,relatime,user_xattr,acl,barrier=1,data=ordered 0 0
 proc /proc proc rw,relatime 0 0
 sysfs /sys sysfs rw,relatime 0 0
 debugfs /sys/kernel/debug debugfs rw,relatime 0 0

 Common.pm won't detect detect that and set boel devstyle to
 standard instead of udev (and thus not mount /a/dev).


 The attached patch extends the regular pattern. But that does not
 really solve the problem. If no-uyok is used, it's not important
 what the running system uses but what the standard kernel will
 require during install.  The new standard kernel will always
 require boel devsytle=udev (or similar) to have /a/dev mounted
 before systemconfigurator will be called from within the
 master.script.


 Best,

 thomas

 --
 Xperia(TM) PLAY
 It's a major breakthrough. An authentic gaming
 smartphone on the nation's most reliable network.
 And it wants your games.
 http://p.sf.net/sfu/verizon-sfdev
 ___
 sisuite-users mailing list
 sisuite-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/sisuite-users



--
Xperia(TM) PLAY
It's a major breakthrough. An authentic gaming
smartphone on the nation's most reliable network.
And it wants your games.
http://p.sf.net/sfu/verizon-sfdev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] systemimager question regarding bad initrd. (still hung)

Hi Olivier:

On Thu, Apr 7, 2011 at 2:04 AM, Olivier LAHAYE olivier.lah...@cea.fr wrote:

 Do you think this problem is solved in latest svn update?

I doubt it, but you could try.

  The problem is that after deploying the image without any trouble, the
  node fails to reboot (kernel panic, no disk found).
  The cause is that the node initrd.img lacks the ahci.ko module (and
  uselesly includes cciss.ko).

Did you actually extract the SystemConfigurator generated initrd and
confirm that it does not include the ahci module?  Some times it's
included but just not called/referenced.

If your headnode has the same hardware as the client nodes, try
copying /etc/modprobe.conf from the headnode to the image, re-image
and reboot, and see if it works.  If the hardware is not identical,
then boot the client in a rescue CD, get a copy of the modprobe.conf
from /tmp and then copy that to your image.

UYOK is probably your best bet though -- why are you hesitant to use
it?  It provides the *best* compatibility between SystemImager
kernel/initrd and deployed OS kernel/initrd.  The gap between the
bundled SystemImager kernel and kernels in current Linux distributions
should be quite huge now...

P.S. Would be good if you could test my RPMs and integrate that with
OSCAR and see if youcome up with any issues.

Cheers,

Bernard

--
Xperia(TM) PLAY
It's a major breakthrough. An authentic gaming
smartphone on the nation's most reliable network.
And it wants your games.
http://p.sf.net/sfu/verizon-sfdev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Help test latest snapshot of SystemImager

Hi all:

On Wed, Apr 6, 2011 at 9:34 PM, Bernard Li bern...@vanhpc.org wrote:

 I forgot to upload the svn4537.ext4 tarball, will do that tomorrow.

The tarball is there now, for those who wanted to build it themselves:

http://download.systemimager.org/~bli/systemimager/

Please report back any issues -- thanks!

Bernard

--
Xperia(TM) PLAY
It's a major breakthrough. An authentic gaming
smartphone on the nation's most reliable network.
And it wants your games.
http://p.sf.net/sfu/verizon-sfdev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Help test latest snapshot of SystemImager

Hi Thomas:

On Thu, Apr 7, 2011 at 2:30 PM, Thomas Zeiser
thomas.zei...@rrze.uni-erlangen.de wrote:

 However, I'm not (yet) really happy that the argument --update-script
 yes|no now is mandatory for si_getimage. (If no --update-script
 argument is given, you get an warning/error in line 200).
 But that's not really a show stopper and I'll adapt my workflow.

That's probably because I took out the unless clause for empty string.
 Try this patch and see if it works:

Index: sbin/si_getimage
===
--- sbin/si_getimage(revision 4539)
+++ sbin/si_getimage(working copy)
@@ -197,10 +197,12 @@
 }

 # be sure $update_script was passed a proper option
-unless(
-   ($update_script eq yes)
-or ($update_script eq no )
-) { die \n$program_name: --update-script must be yes or no.\n$get_help\n\n; }
+if ($update_script) {
+  unless(
+ ($update_script eq yes)
+  or ($update_script eq no )
+  ) { die \n$program_name: --update-script must be yes or
no.\n$get_help\n\n; }
+}

 SystemImager::Server-validate_ip_assignment_option($ip_assignment_method);
 SystemImager::Server-validate_post_install_option($post_install);

Cheers,

Bernard

--
Xperia(TM) PLAY
It's a major breakthrough. An authentic gaming
smartphone on the nation's most reliable network.
And it wants your games.
http://p.sf.net/sfu/verizon-sfdev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Help test latest snapshot of SystemImager

This seems to be a better patch, as it also catches if the user
specified --update-script '' on the CLI:

Index: sbin/si_getimage
===
--- sbin/si_getimage(revision 4539)
+++ sbin/si_getimage(working copy)
@@ -197,10 +197,12 @@
 }

 # be sure $update_script was passed a proper option
-unless(
-   ($update_script eq yes)
-or ($update_script eq no )
-) { die \n$program_name: --update-script must be yes or no.\n$get_help\n\n; }
+if (defined $update_script) {
+  unless(
+ ($update_script eq yes)
+  or ($update_script eq no )
+  ) { die \n$program_name: --update-script must be yes or
no.\n$get_help\n\n; }
+}

 SystemImager::Server-validate_ip_assignment_option($ip_assignment_method);
 SystemImager::Server-validate_post_install_option($post_install);

Cheers,

Bernard

On Thu, Apr 7, 2011 at 2:55 PM, Bernard Li bern...@vanhpc.org wrote:
 Hi Thomas:

 On Thu, Apr 7, 2011 at 2:30 PM, Thomas Zeiser
 thomas.zei...@rrze.uni-erlangen.de wrote:

 However, I'm not (yet) really happy that the argument --update-script
 yes|no now is mandatory for si_getimage. (If no --update-script
 argument is given, you get an warning/error in line 200).
 But that's not really a show stopper and I'll adapt my workflow.

 That's probably because I took out the unless clause for empty string.
  Try this patch and see if it works:

 Index: sbin/si_getimage
 ===
 --- sbin/si_getimage    (revision 4539)
 +++ sbin/si_getimage    (working copy)
 @@ -197,10 +197,12 @@
  }

  # be sure $update_script was passed a proper option
 -unless(
 -       ($update_script eq yes)
 -    or ($update_script eq no )
 -) { die \n$program_name: --update-script must be yes or 
 no.\n$get_help\n\n; }
 +if ($update_script) {
 +  unless(
 +         ($update_script eq yes)
 +      or ($update_script eq no )
 +  ) { die \n$program_name: --update-script must be yes or
 no.\n$get_help\n\n; }
 +}

  SystemImager::Server-validate_ip_assignment_option($ip_assignment_method);
  SystemImager::Server-validate_post_install_option($post_install);

 Cheers,

 Bernard


--
Xperia(TM) PLAY
It's a major breakthrough. An authentic gaming
smartphone on the nation's most reliable network.
And it wants your games.
http://p.sf.net/sfu/verizon-sfdev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] images folder

Hello:

On Thu, Apr 7, 2011 at 7:22 PM, E-Blokos in...@e-blokos.com wrote:

 sorry for my stupid question (don't remember the systemimager options since
 years now)
 is it possible to install the images folder in other part of the filesystem
 ?
 my /etc partition is too small

Images should be stored in /var/lib/systemimager/images.  Are you
saying your /var partition is too small?

You could always create symlinks to other partitions.

Cheers,

Bernard

--
Xperia(TM) PLAY
It's a major breakthrough. An authentic gaming
smartphone on the nation's most reliable network.
And it wants your games.
http://p.sf.net/sfu/verizon-sfdev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Help test latest snapshot of SystemImager

2011-04-06 Thread Bernard Li

Hi Thomas:

On Wed, Apr 6, 2011 at 1:10 PM, Thomas Zeiser
thomas.zei...@rrze.uni-erlangen.de wrote:

 I did not try to rebuild the RPMs as yours cleanly installed (and
 rebuilding requires downloading/extracting lots of packages which
 wont fit into my toy VM environment).

Good to know, thanks!

 I'm still fighting ... Looks like the e2fsprogs included in
 boel_binaries.tar.gz are too old for the generated master script:
 - mke2fs seems to be version 1.40.2
 - script.master contains something like mke2fs -q -t ext4 ${DISK0}2
 - however, mke2fs only learned the -t fstype option in 1.41.0 (if
  I read the e2fsprogs revision history correctly)
 Therefore the execution of mke2fs -t ext4 results in the
 missleading error message invalid blocks count.

 Iff the -t XXX option is removed from the mke2fs call, at least
 creating/mounting the filesystems works and data is rsynced. But I
 still have problems very close to the end, i.e. after the message
 Detecting systemconfigurator:. I have to investigate these, once
 I have some more time.

 (just for completeness: I'm using the standard kernel/initrd)

Thanks for the report, I will backport the following change to my
4.2.x-ext4 tree:

http://trac.systemimager.org/changeset/4490

Hopefully that will solve the problem.  Will cut another snapshot
after I review your other patches and check them in.

Cheers,

Bernard

--
Xperia(TM) PLAY
It's a major breakthrough. An authentic gaming
smartphone on the nation's most reliable network.
And it wants your games.
http://p.sf.net/sfu/verizon-sfdev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Help test latest snapshot of SystemImager

2011-04-06 Thread Bernard Li

Hi Thomas:

I have committed your patches and also updated e2fsprogs to the newer
version, it should work this time.  Please try them out and let me
know if you run into any other issues:

http://download.systemimager.org/~bli/systemimager/

I guess after you've confirmed that ext4 support is working, I could
check that code into the tree.  Would be nice if folks could test on
RHEL6 or Debian/Ubuntu though...  any takers??

Thanks,

Bernard

On Wed, Apr 6, 2011 at 1:22 PM, Thomas Zeiser
thomas.zei...@rrze.uni-erlangen.de wrote:
 On Wed, Apr 06, 2011 at 10:10:06PM +0200, Thomas Zeiser wrote:

 Iff the -t XXX option is removed from the mke2fs call, at least
 creating/mounting the filesystems works and data is rsynced. But I
 still have problems very close to the end, i.e. after the message
 Detecting systemconfigurator:. I have to investigate these, once
 I have some more time.

 Urrrg. For the issure in the final configuration step I was caught
 again in my typical trap: I forgot to activate the mount /dev
 /a/dev -o bind line iin the master script ...

 Thus, without the -t ext4 option and activated bind-mount I
 managed to image an OpenSuSE node with ext4 file system.


 Best,

 thomas


--
Xperia(TM) PLAY
It's a major breakthrough. An authentic gaming
smartphone on the nation's most reliable network.
And it wants your games.
http://p.sf.net/sfu/verizon-sfdev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Help test latest snapshot of SystemImager

2011-04-06 Thread Bernard Li

Hi Franck:

On Wed, Apr 6, 2011 at 8:34 PM, E-Blokos in...@e-blokos.com wrote:

 do you think it will work on the old fedora 10 ?

I don't see why not.  In case you're having trouble installing them, I
could try rebuilding on CentOS 4, or something.

I forgot to upload the svn4537.ext4 tarball, will do that tomorrow.

Thanks,

Bernard

--
Xperia(TM) PLAY
It's a major breakthrough. An authentic gaming
smartphone on the nation's most reliable network.
And it wants your games.
http://p.sf.net/sfu/verizon-sfdev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

[sisuite-users] Fwd: Help test latest snapshot of SystemImager

2011-04-05 Thread Bernard Li

Hi Holger:

[Forwarding this back to the mailing-list]

The mklibs.bef mentioned in your email is only available in trunk and
is not included in the 4.2.x branch that I am working on.  This is the
commit in question:

http://trac.systemimager.org/browser/trunk/initrd_source/mklibs.bef?rev=4495

Perhaps Brian could comment more specifically on your concerns.

Cheers,

Bernard

-- Forwarded message --
From: Dr. Holger Obermaier holger.oberma...@kit.edu
Date: Tue, Mar 29, 2011 at 1:30 AM
Subject: Re: [sisuite-users] Help test latest snapshot of SystemImager
To: Bernard Li bern...@vanhpc.org


Am 26.03.2011 00:24, schrieb Bernard Li:
 Hi all:

 It has been a while since we've made a release (3 years, in fact!) so
 I'm happy to let you guys know that we are working on releasing 4.2.0
 (stable) to the world soon.

Hello Berard,

I had a closer look at your script mklibs.bef which replaces the binary
program from Debian/Ubuntu. I think I found some problems with the
script, which I want to share with you:
 * You are using perls build in function copy to copy the destination.
But copy does not keep the permissions but instead uses the umask. So
for example libc has execution permissions at the source but no
execution permissions at the target.
 * You use a two step method to find the necessary libraries. In the
first step you search for the libraries in the second you replace them
by the preferred libraries. But if you do not find a library in the
first step because the build system does not have this library, there is
nothing to replace. This happened for me with libparted, which is not
installed in the build system but gets generated with parted. So I think
the ldd used to find the libraries should use an LD_LIBRARY_PATH which
includes the user provided library directories. Then there is also no
need for step two any more.
 * The perl method make_path is very new and not available on Scintific
Linux 5.5 and SuSE Enterprise Server 10 SP1 so mkpath may be more portable.

I attached my changes to this mail. Maybe they will be helpful for you.

An other problem I had no solution until now is, that libraries themself
can depend on libraries which is not recognized by the script...

Greetings Holger

--
/*
   Karlsruher Institut für Technologie (KIT)
   Steinbuch Centre for Computing (SCC)
   Abteilung Scientific Computing Services (SCS)
   Dr. Holger Obermaier
   Zirkel 2, Gebäude 20.10, Raum 208
   D-76131 Karlsruhe
   Germany

   phone: +49 (721) 608 - 44862
   web: http://www.scc.kit.edu
   e-mail: holger.oberma...@kit.edu

   KIT - Universität des Landes Baden-Württemberg und nationales
Großforschungszentrum in der Helmholtz-Gemeinschaft
*/
#!/usr/bin/perl -w

# $Id: mklibs.bef 1871 2011-03-28 17:39:36Z rz119 $
# $Date: 2011-03-28 19:39:36 +0200 (Mo, 28. Mär 2011) $
# $Author: rz119 $
# $Revision: 1871 $
# $URL: 
svn+ssh://rz...@icdev-team.rz.uni-karlsruhe.de/srv/svn/OpenOb/scripts/usr/src/packages/SOURCES/systemimager/tools/mklibs.bef
 $

#
# Copyright (C) 2010 Brian Elliott Finley
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; version 2 of the License.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
# General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
#


#   2011.03.23  Brian Elliott Finley
#   * Handle preferred library selection when multiple versions are
# available.

use strict;
use Getopt::Long;
use File::Path qw(mkpath);
use File::Basename;

$ENV{PATH} = /usr/local/sbin:/usr/sbin:/sbin:/usr/local/bin:/usr/bin:/bin;


my $L;
my $v;
my $d;
my $help_info = 
Usage: mklibs.bef -L LIBRARY_PATH -d DEST_LIB_DIR BINARY [BINARY...]
(library path is searched in order specified with -L)
;

GetOptions(
L=s   = \$L,
v = \$v,
d=s   = \$d,
) || die $help_info;

unless( (defined $L) and (defined $d) ) {
print $help_info\n;
exit 1;
}

my @binaries = @ARGV;
my @paths = split(/:/, $L);

if(defined $v) {
print mklibs.bef:\n;
print 
\n;
print  library path (-L):\n;
foreach(@paths) {
print   $_\n;
}
print \n verbose (-v) /* 1=yes, 0=no */:\n  $v\n;
print \n destination (-d):\n  $d\n;
print \n binaries:\n;
}

my %finders;
mkpath($d);
foreach $b (@binaries) {
print   $b\n if(defined $v);
# Search libraries in the libs direcoties provieded by the user
my $cmd = LD_LIBRARY_PATH=\$L\ ldd $b;
open(INPUT,$cmd|) or die;
while(INPUT) {
if(m# = /#) {
my($junk_a, $lib

Re: [sisuite-users] Help test latest snapshot of SystemImager

2011-04-05 Thread Bernard Li

Hi Thomas:

On Tue, Apr 5, 2011 at 12:41 PM, Thomas Zeiser
thomas.zei...@rrze.uni-erlangen.de wrote:

 attached are three small patches to improve compatibility with
 OpenSuSE 11.3:
 - allow parted version numbers a.b in addition to a.b.c
 - avoid non-initialized variable in si_getimage
 - detect date format of OpenSuSE 11.3 kernels for UYOK

Thanks for your patches.

Regarding the change to si_getimage, perhaps this is a better patch?

Index: sbin/si_getimage
===
--- sbin/si_getimage(revision 4535)
+++ sbin/si_getimage(working copy)
@@ -136,7 +136,9 @@
 quiet = \$quiet
 ) || die $help_info;

-$update_script = lc $update_script;
+if ($update_script) {
+  $update_script = lc $update_script;
+}

 #if requested, print help information
 if ($help) {
@@ -196,8 +198,7 @@

 # be sure $update_script was passed a proper option
 unless(
-   ($update_script eq)
-or ($update_script eq yes)
+   ($update_script eq yes)
 or ($update_script eq no )
 ) { die \n$program_name: --update-script must be yes or no.\n$get_help\n\n; }

I do not like setting $update_script to no since that is not the
default (it is perfectly fine not to define $update_script).

I also disallowed $update_script to be  since that doesn't make
sense -- Brian?

Cheers,

Bernard

--
Xperia(TM) PLAY
It's a major breakthrough. An authentic gaming
smartphone on the nation's most reliable network.
And it wants your games.
http://p.sf.net/sfu/verizon-sfdev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Help test latest snapshot of SystemImager

2011-04-05 Thread Bernard Li

Sorry, hit the send button too quickly.  Were you able to test on
OpenSUSE 11.3?  Were you able to build RPMs?  Were you able to test
ext4 support?

Thanks,

Bernard

On Tue, Apr 5, 2011 at 5:16 PM, Bernard Li bern...@vanhpc.org wrote:
 Hi Thomas:

 On Tue, Apr 5, 2011 at 12:41 PM, Thomas Zeiser
 thomas.zei...@rrze.uni-erlangen.de wrote:

 attached are three small patches to improve compatibility with
 OpenSuSE 11.3:
 - allow parted version numbers a.b in addition to a.b.c
 - avoid non-initialized variable in si_getimage
 - detect date format of OpenSuSE 11.3 kernels for UYOK

 Thanks for your patches.

 Regarding the change to si_getimage, perhaps this is a better patch?

 Index: sbin/si_getimage
 ===
 --- sbin/si_getimage    (revision 4535)
 +++ sbin/si_getimage    (working copy)
 @@ -136,7 +136,9 @@
     quiet             = \$quiet
  ) || die $help_info;

 -$update_script = lc $update_script;
 +if ($update_script) {
 +  $update_script = lc $update_script;
 +}

  #if requested, print help information
  if ($help) {
 @@ -196,8 +198,7 @@

  # be sure $update_script was passed a proper option
  unless(
 -       ($update_script eq    )
 -    or ($update_script eq yes)
 +       ($update_script eq yes)
     or ($update_script eq no )
  ) { die \n$program_name: --update-script must be yes or 
 no.\n$get_help\n\n; }

 I do not like setting $update_script to no since that is not the
 default (it is perfectly fine not to define $update_script).

 I also disallowed $update_script to be  since that doesn't make
 sense -- Brian?

 Cheers,

 Bernard


--
Xperia(TM) PLAY
It's a major breakthrough. An authentic gaming
smartphone on the nation's most reliable network.
And it wants your games.
http://p.sf.net/sfu/verizon-sfdev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

[sisuite-users] Help test latest snapshot of SystemImager

2011-03-25 Thread Bernard Li

Hi all:

It has been a while since we've made a release (3 years, in fact!) so
I'm happy to let you guys know that we are working on releasing 4.2.0
(stable) to the world soon.

Here's the current changelog:

4.2.0

- clarify error message with respect to tftp client not found while running
  si_mkbootserver
- added --numeric-ids option to rsync call to work around LDAP bug when username
  is being looked up inside a chroot
- added -fno-tree-scev-cprop to KBUILD_CFLAGS to work around a compile
issue with
  gcc 4.3
- include /var/lib/nfs/statd/sm and /var/lib/nfs/statd/sm.bak in the image
- make --script parameter optional for si_addclients
- run systemconfigurator in verbose mode when setting up an autoinstall client
- use ioctl() to identify IP address of interface (more portable)
- reduce the interval of re-requests made by the BitTorrent seeder on the image
  server to the tracker to 5 minutes
- skip broken symlinks when generating minimal libraries for initrd
- new udev + hotplug-ng hardware detection layer
- added web based monitor GUI for clients.xml
- added /var/www/systemimager directory for all SI web pages
- implement the si_addclients functionality to add /etc/hosts entries in
  si_clusterconfig
- allow si_prepareclient to be executed via ssh by explicitly closing rsync
  stdin
- added support for relatime mount option in BusyBox 1.9.1
- added monitoring and BitTorrent sections to local.cfg example
- added --config option documentation in si_mkbootpackage manpage
- better describe the function of local.cfg in si_mkbootpackage documentation
- enable e1000e network driver by default for i386 and x86_64 kernels
- print out URL to Wiki documentation when BitTorrent binaries are not found
  during installation
- added option --checksum (-c) to si_pcp and si_pushoverrides
- fix warning in si_mkdhcpserver on 64-bit architectures
- use a different DB in si_monitortk if --db is specified
- do not backup tftp file in the xinetd configuration directory when
  si_mkbootserver is run
- updated BOEL kernel to 2.6.31.3 (for i386 and x86_64)
- updated LVM to 2.2.02.53
- removed devmapper
- fixed issue with pvcreate not seeing storage targets with newer kernels
- updated tar to 1.22
- updated kexec-tools to 2.0.1
- added ncurses-devel to RPM spec file BuildRequires
- update xfsprogs to 3.1.4

I have been able to successfully build on RHEL5 and RHEL6, but could
really use your help in testing the code.  Note this is *development*
code, so please do not use in production systems -- USE AT YOUR OWN
RISK :-)

You can find the tarballs and RPMs here:

http://download.systemimager.org/~bli/systemimager/

The snapshot is named 4.1.99 such that when it's ready to be released
it will be renamed 4.2.0.  You'll also notice that there are a bunch
of files labelled ext4 -- these should provide support for ext4
filesystems but have not been tested.  It would be great if you could
try it (eg. on RHEL6).

Would appreciate it if the community could help with testing this
upcoming release and provide feedback in this thread.  It could help
speed up the release process.

Thanks,

Bernard

--
Enable your software for Intel(R) Active Management Technology to meet the
growing manageability and security demands of your customers. Businesses
are taking advantage of Intel(R) vPro (TM) technology - will your software 
be a part of the solution? Download the Intel(R) Manageability Checker 
today! http://p.sf.net/sfu/intel-dev2devmar
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Systemimager-4.2.0 rpm done.

2011-02-23 Thread Bernard Li

Hi Olivier:

On Wed, Feb 23, 2011 at 1:13 AM, Olivier LAHAYE olivier.lah...@cea.fr wrote:

 You're right, this is strange that util-linux was not installed.
 I was building on CentOS-5.5

How did you install CentOS?  Is it with SystemImager?  Is it reproducible?

Cheers,

Bernard

--
Free Software Download: Index, Search  Analyze Logs and other IT data in 
Real-Time with Splunk. Collect, index and harness all the fast moving IT data 
generated by your applications, servers and devices whether physical, virtual
or in the cloud. Deliver compliance at lower cost and gain new business 
insights. http://p.sf.net/sfu/splunk-dev2dev 
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Systemimager-4.2.0 rpm done.

2011-02-17 Thread Bernard Li

Hi Olivier:

On Wed, Feb 16, 2011 at 3:38 AM, Olivier LAHAYE olivier.lah...@cea.fr wrote:

 Do you think the current 4.2.0 is the final 4.2.0 or a pre-release (a RC?)

So far I've only created the 4.2.x branch, I haven't released anything
yet.  Will likely release some snapshots like the following prior to
release:

http://therealms.org/oss/systemimager/snapshots/

When it's released, it will be the final version.  Not sure if it'll
be called 4.2.0 or 4.2.1 yet though...

 The only thing I did change is that I added a BuildRequire: util-linux so
 /sbin/mkfs.cramfs is installed when building the package. (it is required
 during the build)

I could add that, but util-linux should be installed even on minimal
systems since it's needed by most other packages.  What OS are you
building on and how are you constructing your build system such that
util-linux was not needed?

Cheers,

Bernard

--
The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE:
Pinpoint memory and threading errors before they happen.
Find and fix more than 250 security defects in the development cycle.
Locate bottlenecks in serial and parallel code that limit performance.
http://p.sf.net/sfu/intel-dev2devfeb
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] SystemImager BoF at LISA '10 San Jose, CA, USA

2010-11-08 Thread Bernard Li

Hi all:

We're in:

http://www.usenix.org/events/lisa10/bofs.html#sysim

Hope to see you tomorrow evening!

Cheers,

Bernard

On Sat, Nov 6, 2010 at 11:06 PM, Bernard Li bern...@vanhpc.org wrote:
 Hi all:

 We are trying to sign up for a BoF at LISA '10 San Jose:

 http://www.usenix.org/events/lisa10/bofs.html

 For those of you who will be attending the conference, or are local to
 the SF Bay Area, you might want to mark down November 9th,
 9:00-10:00pm down in your calendar because that's the time slot we
 have requested.  Will update this post once we get confirmation.

 This is kind of last minute because I did not know that Brian Elliott
 Finley will be in town for the conference.  Since he's already here,
 why not have a BoF? :)

 Can I please get a quick show of hands from those who will be
 interested in attending?

 Thanks!

 Bernard


--
The Next 800 Companies to Lead America's Growth: New Video Whitepaper
David G. Thomson, author of the best-selling book Blueprint to a 
Billion shares his insights and actions to help propel your 
business during the next growth cycle. Listen Now!
http://p.sf.net/sfu/SAP-dev2dev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

[sisuite-users] SystemImager BoF at LISA '10 San Jose, CA, USA

2010-11-07 Thread Bernard Li

Hi all:

We are trying to sign up for a BoF at LISA '10 San Jose:

http://www.usenix.org/events/lisa10/bofs.html

For those of you who will be attending the conference, or are local to
the SF Bay Area, you might want to mark down November 9th,
9:00-10:00pm down in your calendar because that's the time slot we
have requested.  Will update this post once we get confirmation.

This is kind of last minute because I did not know that Brian Elliott
Finley will be in town for the conference.  Since he's already here,
why not have a BoF? :)

Can I please get a quick show of hands from those who will be
interested in attending?

Thanks!

Bernard

--
The Next 800 Companies to Lead America's Growth: New Video Whitepaper
David G. Thomson, author of the best-selling book Blueprint to a 
Billion shares his insights and actions to help propel your 
business during the next growth cycle. Listen Now!
http://p.sf.net/sfu/SAP-dev2dev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

[sisuite-users] Request for help: Fighting spam on wiki.systemimager.org

2010-11-04 Thread Bernard Li

Hi all:

Lately our Wiki page http://wiki.systemimager.org has been getting a
lot of spam.  I am reaching out to the community to see if anybody has
experience in fighting spam on MediaWiki installations and also to see
if anybody would like to volunteer to be moderator for the Wiki and to
undo spam edits on sight.

Please feel free to reply back to this thread and/or email me off list.

Thanks!

Bernard

--
The Next 800 Companies to Lead America's Growth: New Video Whitepaper
David G. Thomson, author of the best-selling book Blueprint to a 
Billion shares his insights and actions to help propel your 
business during the next growth cycle. Listen Now!
http://p.sf.net/sfu/SAP-dev2dev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Request for help: Fighting spam on wiki.systemimager.org

2010-11-04 Thread Bernard Li

Hi Brian:

On Thu, Nov 4, 2010 at 11:29 AM, Brian Elliott Finley
br...@thefinleys.com wrote:

 How many valid contributors to content on the wiki do we actually have?

That's a good question.  It looks like we have 500+ users:

http://wiki.systemimager.org/index.php/Special:Listusers

Most of them are probably spambots.  Maybe we should just start from
scratch deleting all users and have people re-create their accounts
after implementing captcha.

 Would the use of a captcha plugin do the trick?

Yes, it would probably cut down on the amount of link spamming.

Thanks,

Bernard

--
The Next 800 Companies to Lead America's Growth: New Video Whitepaper
David G. Thomson, author of the best-selling book Blueprint to a 
Billion shares his insights and actions to help propel your 
business during the next growth cycle. Listen Now!
http://p.sf.net/sfu/SAP-dev2dev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Request for help: Fighting spam on wiki.systemimager.org

2010-11-04 Thread Bernard Li

Hi Greg:

On Thu, Nov 4, 2010 at 12:51 PM, Fiumara, Gregory
gregory.fium...@nist.gov wrote:

 I'd second that, and/or require moderator approval for newly created accounts.

Are you referring to moderator approval for new accounts creation
and/or wiki edits?

I'm afraid either way may impose additional overhead that might hinder
active community involvement...  unless of course if captcha does not
do its job well...  does anybody know in practice whether captcha is
sufficient in cutting down spam for a small MediaWIki installation?

Thanks,

Bernard

--
The Next 800 Companies to Lead America's Growth: New Video Whitepaper
David G. Thomson, author of the best-selling book Blueprint to a 
Billion shares his insights and actions to help propel your 
business during the next growth cycle. Listen Now!
http://p.sf.net/sfu/SAP-dev2dev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Ping? Any support? Hello

2010-10-27 Thread Bernard Li

Hi Robby:

On Wed, Oct 27, 2010 at 7:25 AM, Robby Desmond
rdesm...@press.uchicago.edu wrote:

 1) The Wiki is spammed.

Fixed.

 2) Trac is spammed.

This was my bad.  Last year I tried to make Trac available without a
login so that users can submit tickets without having to create an
account -- bad idea.  I'll clean it up soon.

 3) Last stable release (which I'm trying to use, production
 environment and all) is almost 3 years old.
 4) Trac roadmap indicates a version 4.2.x and 5.x in the sights, but
 4.1.7 is still not considered the active stable. Is anyone at the
 project going to clarify the release cycle such as it is?

For the last few years I have been telling people that 4.1.7 is a
stable release even though it is marked unstable.  So if you trust
me, then please use that instead of the really old version. :-)
However, I am still not guaranteeing that your server will not go up
in flames if you use 4.1.7, however I have used it in the past and I'm
sure many people have used it in stable production environments as
well.

 5) Only the developer list seems to get any response from anyone
 involved with SISuite, leading me to believe that support is considered
 beneath the project managers.

I wouldn't say that, necessarily.  But I agree that the traffic for
the users list has been quite low.

This is a call for help.  If you are still using SystemImager (and I
know you are out there!), please try to help others!  We all have our
own jobs and other projects that we work on but if you use the
software and wants it to continue growing, then help out!  Every
little bit helps.

We will be migrating the SystemImager repository to GitHub very soon.
Hopefully once we're there we can kickstart development by attracting
new developers to the project.  In the mean time, I would recommend
you take a look at the SystemImager fork SALI:
https://subtrac.sara.nl/oss/sali.  Once we're on GitHub, the SALI
folks will be creating a fork of our main repository which will make
the changes between the two code base much clearer.

Also, if anybody has experience maintaining MediaWiki and would like
to help us moderate pages for potential spam, please reply back to
this email or send me a private email.

Thanks for your continual support for the SystemImager project.

Best,

Bernard

--
Nokia and ATT present the 2010 Calling All Innovators-North America contest
Create new apps  games for the Nokia N8 for consumers in  U.S. and Canada
$10 million total in prizes - $4M cash, 500 devices, nearly $6M in marketing
Develop with Nokia Qt SDK, Web Runtime, or Java and Publish to Ovi Store 
http://p.sf.net/sfu/nokia-dev2dev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Client not getting the base image

2010-09-24 Thread Bernard Li

Have you started the flamethrower daemon on the server?  If you are a
new SystemImager user I would recommend you try the rsync transport
instead of advanced options like flamethrower or BitTorrent.

Cheers,

Bernard

On Fri, Sep 24, 2010 at 9:36 AM, Dheeraj KV kvdhee...@indiatimes.com wrote:
 Hi
         Can anyone help me out in this case. I m really struggling with this 
 issue. Client is not getting the image it seems. The following is the output 
 on the terminal

 ###
 EXT3 FS on cciss/c0d0p2, internal journal
 EXT3-fs: mounted filesystem with ordered data mode.
 mke2fs -q -j /dev/cciss/c0d0p1 || shellout
 tune2fs -L /boot /dev/cciss/c0d0p1
 tune2fs 1.38 (30-Jun-2005)
 mkdir -p /a/boot || shellout
 mount /dev/cciss/c0d0p1 /a/boot -t ext3 -o defaults || shellout
 kjournald starting.  Commit interval 5 seconds
 EXT3 FS on cciss/c0d0p1, internal journal
 EXT3-fs: mounted filesystem with ordered data mode.
 mkdir -p /a/proc || shellout
 mount proc /a/proc -t proc -o defaults || shellout
 mkdir -p /a/sys || shellout
 mount sysfs /a/sys -t sysfs -o defaults || shellout

 flamethrower_client(base4_image) tarpipe
 -
 udp-receiver --interface eth2 --portbase 9006 --nokbd --nosync | tar 
 --overwrite -xp -f - -C /a
 Udp-receiver 2006-03-25
 UDP receiver for (stdout) at 172.31.0.19 on eth2
 received message, cap=0009
 Connected as #0 to 172.31.24.151
 Listening to multicast on 236.31.24.151
 /etc/init.d/rcS: /etc/init.d/rcS: 148: echo: Argument list too long
 Dropped by server now=2361 last=2358
 Transfer complete.
 tar: Read 2048 bytes from -
 tar: Unexpected EOF in archive
 tar: Unexpected EOF in archive
 tar: Error is not recoverable: exiting now
 flamethrower_client: Proceeding with retry 1 of 7
 ...udp-receiver --interface eth2 --portbase 9006 --nokbd --nosync | tar 
 --overwrite -xp -f - -f - -C /a
 Udp-receiver 2006-03-25
 UDP receiver for (stdout) at 172.31.0.19 on eth2
 tar: Multiple archive files require `-M' option
 Try `tar --help' or `tar --usage' for more information.
 received message, cap=0009
 Connected as #0 to 172.31.24.151
 Listening to multicast on 236.31.24.151
 flamethrower_client: Proceeding with retry 2 of 7
 ...udp-receiver --interface eth2 --portbase 9006 --nokbd --nosync | tar 
 --overwrite -xp -f - -f - -f - -C /a
 Udp-receiver 2006-03-25
 UDP receiver for (stdout) at 172.31.0.19 on eth2
 tar: Multiple archive files require `-M' option
 Try `tar --help' or `tar --usage' for more information.
 Bad server reply 0006. Other transfer in progress?
 flamethrower_client: Proceeding with retry 3 of 7
 ...udp-receiver --interface eth2 --portbase 9006 --nokbd --nosync | tar 
 --overwrite -xp -f - -f - -f - -f - -C /a
 Udp-receiver 2006-03-25
 UDP receiver for (stdout) at 172.31.0.19 on eth2
 tar: Multiple archive files require `-M' option
 Try `tar --help' or `tar --usage' for more information.
 Bad server reply 0006. Other transfer in progress?
 flamethrower_client: Proceeding with retry 4 of 7
 ...udp-receiver --interface eth2 --portbase 9006 --nokbd --nosync | tar 
 --overwrite -xp -f - -f - -f - -f - -f - -C /a
 Udp-receiver 2006-03-25
 UDP receiver for (stdout) at 172.31.0.19 on eth2
 tar: Multiple archive files require `-M' option
 Try `tar --help' or `tar --usage' for more information.
 Bad server reply 0006. Other transfer in progress?
 flamethrower_client: Proceeding with retry 5 of 7
 ...udp-receiver --interface eth2 --portbase 9006 --nokbd --nosync | tar 
 --overwrite -xp -f - -f - -f - -f - -f - -f - -C /a
 Udp-receiver 2006-03-25
 UDP receiver for (stdout) at 172.31.0.19 on eth2
 tar: Multiple archive files require `-M' option
 Try `tar --help' or `tar --usage' for more information.
 Bad server reply 0006. Other transfer in progress?
 flamethrower_client: Proceeding with retry 6 of 7
 ...udp-receiver --interface eth2 --portbase 9006 --nokbd --nosync | tar 
 --overwrite -xp -f - -f - -f - -f - -f - -f - -f - -C /a
 Udp-receiver 2006-03-25
 UDP receiver for (stdout) at 172.31.0.19 on eth2
 tar: Multiple archive files require `-M' option
 Try `tar --help' or `tar --usage' for more information.
 Bad server reply 0006. Other transfer in progress?
 flamethrower_client: Proceeding with retry 7 of 7
 ...udp-receiver --interface eth2 --portbase 9006 --nokbd --nosync | tar 
 --overwrite -xp -f - -f - -f - -f - -f - -f - -f - -f - -C /a
 Udp-receiver 2006-03-25
 UDP receiver for (stdout) at 172.31.0.19 on eth2
 tar: Multiple archive files require `-M' option
 Try `tar --help' or `tar --usage' for more information.
 Bad server reply 0006. Other transfer in progress?


 BusyBox v1.1.1 (2007.12.06-11:09+) Built-in shell (ash)
 Enter 'help' for a list of built-in commands.

 sh: can't access tty; job control turned off
 /scripts/pre-install #
 flamethrower_client: FATAL: Initial attempt and 7 retries failed!
 Killing off running processes.

 write_variables

[sisuite-users] RFC: Move SystemImager development to GitHub

2010-09-15 Thread Bernard Li

Hi all:

I have recently discussed with Brian Finley the possibility of moving
SystemImager development from SVN hosted at systemimager.org to Git
hosted at GitHub and we both agreed that it's a good idea.

I have also consulted the team from SARA who is currently working on
the SALI fork of SystemImager, and they are also onboard.

The idea would be to have the source code repository hosted at GitHub
owned by user 'systemimager' which will be publicly available and
anybody would then be free to fork the repository and work on
features/enhancements in their own scratch space.  A good example is
the current SALI fork.  If we host it at GitHub, it would make the
changes that were done to mainline SystemImager much clearer/easy to
track and also allow the SystemImager team to cherry pick part or all
of the commits back to the official release.  We will also be able
to get more exposure hosting our project at GitHub.

What do the general community think about this?

Thanks,

Bernard

--
Start uncovering the many advantages of virtual appliances
and start using them to simplify application deployment and
accelerate your shift to cloud computing.
http://p.sf.net/sfu/novell-sfdev2dev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] si_monitor debugging

2010-09-07 Thread Bernard Li

Hi Jeremy:

On Tue, Sep 7, 2010 at 2:50 PM, Jeremy Enos je...@ncsa.uiuc.edu wrote:

 A couple other things:
 I read this item to check from the wiki page:
 Are you sure the clients are using the correct boot parameters? (remember
 that you can check the parameters directly from the clients -if you have the
 access to the real console- looking in/proc/cmdline)
 But the page doesn't mention what the correct boot parameters are though.  I
 remember something like MONITOR_SERVER=xyz or something maybe?
 http://wiki.systemimager.org/index.php/Monitoring
 Also, the page above seems to have gotten infected with paragraphs of spam
 inserted randomly in it.  Might be worth checking out.

This is funny.  The spam edits you mentioned actually overwrote the
information you need.

I just undid the changes.

Cheers,

Bernard

--
This SF.net Dev2Dev email is sponsored by:

Show off your parallel programming skills.
Enter the Intel(R) Threading Challenge 2010.
http://p.sf.net/sfu/intel-thread-sfd
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] grub2 status

2010-09-01 Thread Bernard Li

Hi Chris:

On Wed, Sep 1, 2010 at 10:28 AM, Chris Scheller sche...@pobox.com wrote:

 what is the status of the work to support grub2? i see mention of work
 being done back in march of 2010 but nothing since.

I believe Brian is working on that in trunk.

You can also look at the SALI project:

https://subtrac.sara.nl/oss/sali

Cheers,

Bernard

--
This SF.net Dev2Dev email is sponsored by:

Show off your parallel programming skills.
Enter the Intel(R) Threading Challenge 2010.
http://p.sf.net/sfu/intel-thread-sfd
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Choosing between si_clusterconfig and si_addclients

2010-06-01 Thread Bernard Li

Hi Larry:

On Tue, Jun 1, 2010 at 11:10 AM, Larry Anta la...@ryerson.ca wrote:

 I'm still a bit fuzzy about when it's best to use si_clusterconfig (which
 appears to maintain a relationship between images and clients in a file
 named cluster.xml) and when to use si_addclients (which maintains that
 relationship using soft links in /var/lib/systemimager/scripts).

si_clusterconfig is supposed to replace si_addclients as noted by the
message in this changeset:

http://trac.systemimager.org/changeset/4321/trunk

Cheers,

Bernard

--

___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] rsync problems

2010-03-08 Thread Bernard Li

Hi Sandy:

On Mon, Mar 8, 2010 at 9:03 AM, Alexander James Spence [axs]
a...@aber.ac.uk wrote:

 There is 1G of RAM, each time it complains that tmpfs has run out of space
 when transferring the file list, any clues where to start to fix this
 problem?

Can you please post the exact error message you got?

How big is your image?  And you are using the rsync transport
(default) correct?  You did mention you're getting rsync errors but
the BItTorrent transport also uses rsync for some portion of the
installation.

Also, please post the version of SystemImager you're using and the
distro/version or your image server.

Regards,

Bernard

--
Download Intel#174; Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Updating kernel to 2.6.33

2010-03-04 Thread Bernard Li

Hi Bryan:

On Thu, Mar 4, 2010 at 9:36 AM, Bryan Washer bwas...@vitesse.com wrote:

 I really do not want to update my SystemImager infrastructure as I have
 about 10 of them meshed throughout the organization and that would be quite
 a task, but I have a new load of HP DL160’s that I Need to add support for
 and the IGB driver is not supported in the kernel until 2.6.27, and I have
 had one hell of a time trying to back port the driver so I thought I would
 just update the SI kernel.

Have you tried using UYOK?  That way you shouldn't need to update the
SystemImager kernel:

http://wiki.systemimager.org/index.php/UYOK

Cheers,

Bernard

--
Download Intel#174; Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] LVM2.2.02.57 (mklibs, line 478, in module, raise No library provides non-weak + symbol) ERROR

2010-01-15 Thread Bernard Li

Hi Thierry:

On Fri, Jan 15, 2010 at 2:05 AM, Thierry Iceta thierry.ic...@bull.net wrote:

 My problem is when I try to include new LVM2.2.02.56 in
 systemimager-4.0.2 production environment  and compile (make ) under
 Fedora12
 I got an error when mklibs try to link sbin/* directory where lvm
 binaries are located.
 No probleme with LVM22.02.53 and under
 probleme  with LVM22.02.56 and   This LVM is requested for FEDORA12
 Here under you will find most of informations

Is there a reason why you need that version specifically?  Are you
saying that if you image a client using LVM 22.02.53 it is not
recognizable by Fedora 12?

Cheers,

Bernard

--
Throughout its 18-year history, RSA Conference consistently attracts the
world's best and brightest in the field, creating opportunities for Conference
attendees to learn about information security's most important issues through
interactions with peers, luminaries and emerging and established companies.
http://p.sf.net/sfu/rsaconf-dev2dev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] bnx2 firmware issues during boot.

2009-12-16 Thread Bernard Li

Hi Michael:

On Tue, Dec 15, 2009 at 9:48 PM, Michael Pawlowsky
mi...@clearskymedia.ca wrote:

 I noticed a few other people have had issues with the bnx2 firmware not being 
 loaded when booting from a systemimager bootcd.
 I was wondering if someone could explain the exact work around for this.

 I've tried UYOK with no luck.

 When following this page:  http://wiki.systemimager.org/index.php/UYOK

 I also get --no-rsyncd is not compatible with --server with executing 
 si_prepareclient --server servername --no-rsyncd --my-modules so I've been 
 leaving out the --no-rsyncd.

Which version of SystemImager did you try that with?  Did all versions
give that warning?

 I've tried using insmod to insert the bnx2.ko module into the kernel, but it 
 tells me it is already there.

 I am using version 4.1.7.  (Started with 4.0.2, to 4.1.6 and now 4.1.7)

 4.0.2 had no driver issues for the Broadcom Xtreme but was having issues with 
 LVM.

The 4.1.7 kernel has that module built in.  I'm curious why it
couldn't detect your nic?  Can you run `lspci -v` on that box and
paste the output relevant to your nic?

With these kinds of issues, it might work better if you're on IRC so
that we can troubleshoot this interactively.

Cheers,

Bernard

--
This SF.Net email is sponsored by the Verizon Developer Community
Take advantage of Verizon's best-in-class app development support
A streamlined, 14 day to market process makes app distribution fast and easy
Join now and get one step closer to millions of Verizon customers
http://p.sf.net/sfu/verizon-dev2dev 
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Manual configuration patch

2009-12-15 Thread Bernard Li

Hi Judd:

Thanks for the patch.

Could you please provide a unified diff patch?  SVN diff works too.

Also, is your diff against the code in trunk?  Or a specific version
of SystemImager?

Cheers,

Bernard

On Tue, Dec 15, 2009 at 3:15 PM, Judd Kennedy jkenn...@neoninc.org wrote:
 Sisuite users,



 I have generated a patch to allow manual system configuration from initrd.



 ---



 The gyst:



 * Allows user intervention if DHCP services and variables needed for imaging
 are not available, with the option to *change* a value without rebooting.

 * Will use dialog for menuing, if available.  Xdialog is also an option
 (requires a big, fat, custom initrd)

 * Allows user to enter the following details at image time: HOSTNAME,
 CUSTOMER_HOSTNAME, DOMAINNAME, DEVICE, IPADDR, NETMASK, NETWORK, BROADCAST
 ADDRESS, GATEWAY, GATEWAYDEV, IMAGESERVER, IMAGENAME



 Available at:

 http://juddy.org/functions_interactive.patch



 WJK



 --
 This SF.Net email is sponsored by the Verizon Developer Community
 Take advantage of Verizon's best-in-class app development support
 A streamlined, 14 day to market process makes app distribution fast and easy
 Join now and get one step closer to millions of Verizon customers
 http://p.sf.net/sfu/verizon-dev2dev
 ___
 sisuite-users mailing list
 sisuite-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/sisuite-users



--
This SF.Net email is sponsored by the Verizon Developer Community
Take advantage of Verizon's best-in-class app development support
A streamlined, 14 day to market process makes app distribution fast and easy
Join now and get one step closer to millions of Verizon customers
http://p.sf.net/sfu/verizon-dev2dev 
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] www.systemimager.org currently unavailable

2009-12-11 Thread Bernard Li

Just to let you guys know, the server is back online now.

Thanks,

Bernard

On Tue, Dec 8, 2009 at 12:02 PM, Bernard Li bern...@vanhpc.org wrote:
 Hi all:

 The server hosting our web, svn and trac services is currently being
 relocated to a new data center and thus is unavailable.

 I will update this thread once the server is back up and running.

 Sorry for any inconvenience caused.

 Cheers,

 Bernard


--
Return on Information:
Google Enterprise Search pays you back
Get the facts.
http://p.sf.net/sfu/google-dev2dev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] bnx2 firmware problem

2009-12-11 Thread Bernard Li

And as Olivier pointed out, 4.1.7-pre's stock kernel should also work,
since it has the bnx2 module compiled in.

Good luck,

Bernard

On Fri, Dec 11, 2009 at 9:20 AM, Bernard Li bern...@vanhpc.org wrote:
 Hi Derek:

 On Fri, Dec 11, 2009 at 6:46 AM, Derek Harkness dhark...@umich.edu wrote:

 I'm trying to image a Dell r905 that requires the bnx2 firmware to be loaded 
 by the OS.  I've got the firmware on initrd image and can successful load 
 the firmware into the kernel by hand but hotplug just won't load it when it 
 needs to.

 Does anyone have a solution to this?

 I'm using systemimage 4.1.6 with everything else pretty vanilla.

 Assuming the OS you are trying to provision supports bnx2, give UYOK a
 try and see if that works:

 http://wiki.systemimager.org/index.php/UYOK

 Cheers,

 Bernard


--
Return on Information:
Google Enterprise Search pays you back
Get the facts.
http://p.sf.net/sfu/google-dev2dev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] bnx2 firmware problem

2009-12-11 Thread Bernard Li

Hi Derek:

On Fri, Dec 11, 2009 at 5:50 PM, Derek Harkness dhark...@umich.edu wrote:

 Thanks tried the stand si 4.1.6 kernel and it does work.  It was the
 UYOK kernel and initrd created by prepareclient that didn't work.  The
 problem seems to stem from the fact I'm imaging a debian system that
 doesn't compile the bnx2 firmware into the kernel.  It still seem odd
 that the /etc/hotplug/firmware.agent never gets called.

Which step are you talking about?  During imaging or after imaging is
done and the OS (Debian) is being booted for the first time?

Cheers,

Bernard

--
Return on Information:
Google Enterprise Search pays you back
Get the facts.
http://p.sf.net/sfu/google-dev2dev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

[sisuite-users] www.systemimager.org currently unavailable

2009-12-08 Thread Bernard Li

Hi all:

The server hosting our web, svn and trac services is currently being
relocated to a new data center and thus is unavailable.

I will update this thread once the server is back up and running.

Sorry for any inconvenience caused.

Cheers,

Bernard

--
Return on Information:
Google Enterprise Search pays you back
Get the facts.
http://p.sf.net/sfu/google-dev2dev
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] systemimager 4.1.7 svn 4474: GREAT!

2009-11-25 Thread Bernard Li

Hi Olivier:

Can you please file a bug in Trac http://trac.systemimager.org and
assign it to me (my username is 'bli').

Thanks,

Bernard

On Wed, Nov 25, 2009 at 4:25 AM, Olivier LAHAYE olivier.lah...@cea.fr wrote:

 Oups, just found a small bug. It looks like the autoinstallscript.conf is 
 incompletely generated by oscar. If this is the case, the 
 si_mkautoinstallscript crash instead of printing an  error. Looking at it:

 log from OSCAR svn (oscar-6.0.5r9073-1): just at the end of image generation.

 Image build finished.
 Added Disk Table for oscarimage based on 
 /usr/share/oscar/oscarsamples/cciss.disk
 Running: /usr/sbin/si_mkautoinstallscript -quiet --autodetect-disks -image 
 oscarimage -force -ip-assignment static -post-install reboot ... Use of 
 uninitialized value in numeric eq (==) at 
 /usr/lib/systemimager/perl/SystemImager/Server.pm line 622, $TEMPLATE line 
 165.
 Use of uninitialized value in numeric gt () at 
 /usr/lib/systemimager/perl/SystemImager/Server.pm line 656, $TEMPLATE line 
 165.
 Use of uninitialized value in numeric gt () at 
 /usr/lib/systemimager/perl/SystemImager/Server.pm line 656, $TEMPLATE line 
 165.
 Use of uninitialized value in numeric ge (=) at 
 /usr/lib/systemimager/perl/SystemImager/Server.pm line 909, $TEMPLATE line 
 175.
 done

 Maybe it is related to my disks /dev/cciss/c0d0{,p{1,5,6}} part tags 
 missing in disk section.
 following files are attached:

 /var/lib/systemimager/images/oscarimage/etc/systemimager/autoinstallscript.conf
 /usr/share/oscar/oscarsamples/cciss.disk

 /usr/lib/systemimager/perl/SystemImager/Server.pm:610
 DB27 p Dumper $xml_config-{disk}-{$dev}
 $VAR1 = {
          'label_type' = 'msdos',
          'dev' = '/dev/cciss/c0d0',
          'part' = {},
          'unit_of_measurement' = 'MB'
        };

 'part' is empty = line 612: $highest_p_or_e_part_num is never defined, then 
 boom.

 On oscar side: I need to check how to fix the generation of 
 autoinstallscript.conf, on your side, you need to report no partitions error 
 message or the such.

 Best regards.

 Olivier.

 Le mercredi 25 novembre 2009 10:10:49 Olivier LAHAYE, vous avez écrit :

 Hi Bernard,

 As for releasing 4.1.7, IMHO, for having used it with OSCAR 6.0.5pre and 
 CentOS 5.4, it works fine.
 For other purposes, I can't tell if the code is ready, but I don't see any 
 reasons why it wouldn't.

 For my part, the great advantage of this version is its kernel that supports 
 the HP Raid Array (cciss) without FIFO Full error messages.
 So, for all people wanting to use OSCAR on HP servers that have a cciss raid 
 array, I think it would be very usefull as it would avoid the use your own 
 kernel requirement.

 In simple words yes it would be very helpful to have a new release :-)

 As for grub2: it is the standard bootloader for Mandriva Linux 2010, and it 
 is supported by at least SuSE 11 and debian unstable if I'm correct.
 Booting on GPT partition table is only possible with grub2. grub1 and LILO 
 are unable to boot a GPT partition table.

 My cluster head is equipped with only a big 6To raid array and so far only 
 Mandriva 2010 was able to install itself on this hardware.
 I'm planing to port Mandriva grub2 rpm to CentOS 5.4 so I could image my 
 head and then, using a mandriva live usb key, I would grub2 the boot.

 As for the patch, I'd be glad to create it, unfortunately, if you read 
 oscar-devel mailing list you'll notice that I'm a perl newbee. For now, I'll 
 look at porting grub2 package and then, if I succeed, I'll give a try though.

 Best Regards,

 Olivier.

 Le mercredi 25 novembre 2009 06:20:40 Bernard Li, vous avez écrit :
  Hi Olivier:
 
  On Thu, Nov 19, 2009 at 12:28 AM, Olivier LAHAYE olivier.lah...@cea.fr 
  wrote:
 
   Just to report that I'm testing latest svn version of systemimager 
   (4.1.7 svn 4474) on CentOS 5.4 running on an HP system with a raid array 
   (CCISS) and it is the 1st version that runs flawlessly. No need to run 
   my own kernel to avoid Fifo full errors on cciss decices.
  
   No bug found so far. I'm testing this within the latest svn version of 
   OSCAR (6.0.5 svn 9034)
 
  Thanks for the report!  Recently I have updated the kernel as well as
  LVM and tar.  I am glad that this is working out for you.
 
  I was thinking of making a release when I get the chance.  If you
  think the code is ready, I can make a 4.1.7 release which the OSCAR
  team can then bundle -- would that be helpful?
 
   One question though: does system imager support booting on GPT partition 
   table? In other words: does it support grub2. If not, is it aimed?
 
  Does grub1 (grub-legacy) not support GPT partition table?  I know for
  certain that grub2 is not currently supported, but that does not mean
  it cannot be added in the future.  However, your underlying OS would
  need to be running grub2 as well.
 
  You are more than welcome to work on a patch and submit that to us.
  The code for SystemConfigurator which would be in-charge

Re: [sisuite-users] si_mkautoinstallscript (svn 4473) : Use of uninitialized value Server.pm line 671 and 916

2009-11-25 Thread Bernard Li

Hi Olivier:

On Thu, Nov 12, 2009 at 4:39 AM, Olivier LAHAYE olivier.lah...@cea.fr wrote:

 It helped (no more error), but in the meantime, if there is an error in 
 config file generated by oscar, si_mkautoinstallscript fails with the error 
 below instead of reporting a human readable error like config file not 
 correct at line: device /dev/cciss/c0d06 does not exists. Error.
 It is a little bit like /bin/ls /sdrbhzdjkfbhqdf that would dump core 
 instead of telling /sdrbhzdjkfbhqdf: no such file or directory.

Can you please file a bug in Trac and assign it to me?

Thanks,

Bernard

--
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day 
trial. Simplify your report design, integration and deployment - and focus on 
what you do best, core application coding. Discover what's new with
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] systemimager 4.1.7 svn 4474: GREAT!

2009-11-24 Thread Bernard Li

Hi Olivier:

On Thu, Nov 19, 2009 at 12:28 AM, Olivier LAHAYE olivier.lah...@cea.fr wrote:

 Just to report that I'm testing latest svn version of systemimager (4.1.7 svn 
 4474) on CentOS 5.4 running on an HP system with a raid array (CCISS) and it 
 is the 1st version that runs flawlessly. No need to run my own kernel to 
 avoid Fifo full errors on cciss decices.

 No bug found so far. I'm testing this within the latest svn version of OSCAR 
 (6.0.5 svn 9034)

Thanks for the report!  Recently I have updated the kernel as well as
LVM and tar.  I am glad that this is working out for you.

I was thinking of making a release when I get the chance.  If you
think the code is ready, I can make a 4.1.7 release which the OSCAR
team can then bundle -- would that be helpful?

 One question though: does system imager support booting on GPT partition 
 table? In other words: does it support grub2. If not, is it aimed?

Does grub1 (grub-legacy) not support GPT partition table?  I know for
certain that grub2 is not currently supported, but that does not mean
it cannot be added in the future.  However, your underlying OS would
need to be running grub2 as well.

You are more than welcome to work on a patch and submit that to us.
The code for SystemConfigurator which would be in-charge of the grub
management is available here:

http://svn.systemimager.org/listing.php?repname=systemconfiguratorpath=%2Fsc=0

Cheers,

Bernard

--
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day 
trial. Simplify your report design, integration and deployment - and focus on 
what you do best, core application coding. Discover what's new with
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] SystemImager 4.0.2 and CVE-2008-5156

2009-11-18 Thread Bernard Li

Hi Justin:

Thanks for bringing this to our attention.  I'm at SC this week but
once I'm back I'll have a look at this issue to see whether it is
still valid for our latest released version(s).

Cheers,

Bernard

On Wed, Nov 18, 2009 at 11:26 AM, Justin Moninger
justin.monin...@gmail.com wrote:
 Hi All,



 I’m new to using SystemImager, but a coworker turned me on to it and he
 loves it.  It is the perfect tool for my disaster recovery plan for a system
 I’m working on.  Prior to putting things into production I have to run
 things by our IT security engineers, and one of them found the following
 (details below):





 CVE-2008-5156



 I did look through the SysImager Trac instance, but found nothing related.
 My research led me to believe this is Debian only (I use Red Hat/CentOS),
 but I don’t fully understand the problem.  Does anyone know if this is an
 issue or if it was patched for 4.0.2?  Otherwise is the mitigation just
 chmod’ing /tmp after pushing the image, or is this a build time
 vulnerability?



 Thanks!!!



 Justin

 -

 Vulnerability Summary for CVE-2008-5156

 Original release date:11/18/2008
 Last revised:11/18/2008
 Source: US-CERT/NIST

 Overview

 si_mkbootserver in systemimager-server 3.6.3 allows local users to overwrite
 arbitrary files via a symlink attack on a (1) /tmp/*.inetd.conf or (2)
 /tmp/pxe.conf.*.tmp temporary file.

 Impact

 CVSS Severity (version 2.0):

 CVSS v2 Base Score:6.9 (MEDIUM) (AV:L/AC:M/Au:N/C:C/I:C/A:C) (legend)
 Impact Subscore: 10.0
 Exploitability Subscore: 3.4

 CVSS Version 2 Metrics:

 Access Vector: Locally exploitable
 Access Complexity: Medium
 Authentication: Not required to exploit
 Impact Type:Allows unauthorized disclosure of information; Allows
 unauthorized modification; Allows disruption of service

 --
 Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
 trial. Simplify your report design, integration and deployment - and focus
 on
 what you do best, core application coding. Discover what's new with
 Crystal Reports now.  http://p.sf.net/sfu/bobj-july
 ___
 sisuite-users mailing list
 sisuite-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/sisuite-users



--
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day 
trial. Simplify your report design, integration and deployment - and focus on 
what you do best, core application coding. Discover what's new with
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] si_mkautoinstallscript (svn 4473) : Use of uninitialized value Server.pm line 671 and 916

2009-11-10 Thread Bernard Li

Hi Olivier:

On Tue, Nov 10, 2009 at 7:15 AM, Olivier LAHAYE olivier.lah...@cea.fr wrote:

 So no I must figure out what went wrong with the command that generated  
 /var/lib/systemimager/images/myimage/etc/systemimager/autoinstallscript.conf

 mksidisk -A --name myimage --file /usr/share/oscar/oscarsamples/cciss.disk

 The above command succeed but generates wrong things.
 /var/lib/systemimager/images/myimage/etc/systemconfig/systemconfig.conf has 
 incomplete BOOTDEV = /dev/cciss/c0d0p (missing 1)
 /var/lib/systemimager/images/myimage/etc/systemimager/autoinstallscript.conf

mksidisk belongs to the package systeminstaller-oscar which is
maintained by the OSCAR developers -- I think you're best to contact
them for help.

Typically in an OSCAR installation, the autoinstallscript.conf for
your image is generated by SystemInstaller, and SystemImager generates
the master installation script using that as a template.

I noticed that you have already patched SystemInstaller -- did that not help?

http://www.mail-archive.com/oscar-de...@lists.sourceforge.net/msg08990.html

Cheers,

Bernard

--
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day 
trial. Simplify your report design, integration and deployment - and focus on 
what you do best, core application coding. Discover what's new with
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Have Two Images, But Can't Get Either One To Install

2009-11-10 Thread Bernard Li

Hi Kyle:

Are you using UYOK?

Also I would recommend giving 4.1.6 a try.

Cheers,

Bernard

On Wed, Nov 4, 2009 at 1:11 PM, Kyle Wagner kyman2...@gmail.com wrote:
 Hey guys, I'm new to SystemImager and I'm having a little bit of trouble...
 Here's my setup:  I have a headnode set up running Linux Fedora 11 64bit and
 it has a dhcp server assigning IP addresses to 4 cluster nodes on my
 network.  Two of these cluster nodes have Linux Fedora 11 64bit installed,
 various other software, and then SystemImager 4.0.2.  One of my cluster
 nodes was custom partitioned prior to Linux being installed.  The other was
 just partitioned automatically with the Linux installation.  I have
 successfully made an image of both nodes and I'm trying to send them to my
 other two nodes (which have blank hard drives).  Both nodes boot from the
 network properly, connect to my image server (which is also my headnode),
 and find the image just fine.  It's the autoinstallation of the images is
 where the trouble is.

 When systemimager tries to install the image with the custom partitions,
 everything is fine until this point:

 *The partitions are created*
 
 mount sysfs /a/sys -t sysfs -o defaults || shellout
 Quietly installing image...
 df: /a/tmp: No such file or directory
 df: /a/boot: No such file or directory
 df: /a/var: No such file or directory
 df: /a/share: No such file or directory

 These last four lines are just reprinted to the screen over and over until
 the whole process finally fails, saying:

 mount: can't find a/var in /etc/fstab or /etc/mtab



 When systemimager tries to install the image with the automatic Linux
 partitioning, the image seems to install alright.

 rsync -aHS --exclude=lost+found/  --exclude=/proc/*  --numberic-ids
 192.168.0.100: : quadnode/ /a/

 but then I receive something similar to:

 swapper: page allocation failure. order:0, mode:0x20

 Call Trace:
  IRQ  [8020e0ee] __alloc_pages+0x288/0x2a1
  [80233114] tcp_v4_do_rcv+0x26/0x290

  [8024fb45] cache_alloc_refill+0x23f/0x45e
  [80298982] __kmalloc+0x50/0x57
  [80228233] __alloc_skb+0x5a/0x133
  [803a4a5d] nv_alloc_rx_optimized+0x58/0x18c

  [803a6e30] nv_nic_irq_optimized+0x87/0x1d9
  [8020f098] handle_IRQ_event+0x25/0x53
  [80210374] __do_softirq+0x46/0x90
  [8028ab8f] handle_fasteoi_irq+0x5b/0x88

  [8025d66e] do_IRQ+0xd7/0x132
  [8025ba07] default_idle+0x0/0x3a
  [802509f1] ret_from_intr+0x0/0xa
  EOI  [8045e907] udp_poll+0x0/0x128
  [8025ba2d] default_idle+0x26/0x3a

  [8023e163] cpu_idle+0x3d/0x5c
  [805e58e9] start_kernel+0x294/0x2a0
  [805e5140] _sinittext+0x140/0x144

 Mem-info:
 DMA per-cpu:
 CPU0: Hot: hi:0, btch:   1 usd:   0   Cold: hi:0, btch:   1 usd:
 0

 DMA32 per-cpu:
 CPU0: Hot: hi:  186, btch:  31 usd:  30   Cold: hi:   62, btch:  15 usd:
 48
 Active:170237 inactive:288513 dirty:24931 writeback:1 unstable:0
  free:2537 slab:47109 mapped:7013 pagetables:1123 bounce:0

 DMA free:8024kB min:28kB low:32kB high:40kB active:3096kB inactive:0kB
 present:11164kB pages_scanned:128 all_unreclaimable? no
 lowmem_reserve[]: 0 2003 2003
 DMA32 free:2124kB min:5712kB low:7140kB high:8568kB active:677852kB
 inactive:1154052kB present:2051184kB pages_scanned:0 all_unreclaimable? no

 lowmem_reserve[]: 0 0 0
 DMA: 0*4kB 1*8kB 1*16kB 0*32kB 1*64kB 0*128kB 1*256kB 1*512kB 1*1024kB
 1*2048kB 1*4096kB = 8024kB
 DMA32: 1*4kB 1*8kB 0*16kB 0*32kB 1*64kB 0*128kB 0*256kB 0*512kB 0*1024kB
 1*2048kB 0*4096kB = 2124kB

 Swap cache: add 188636, delete 61524, find 121/155, race 0+0
 Free swap  = 5101816kB
 Total swap = 5855208kB
 Free swap:   5101816kB
 524000 pages of RAM
 9148 reserved pages
 177075 pages shared
 127112 pages swap cached

 swapper: page allocation failure. order:0, mode:0x20



 Any help is greatly appreciated! Thanks!

 --
 Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
 trial. Simplify your report design, integration and deployment - and focus
 on
 what you do best, core application coding. Discover what's new with
 Crystal Reports now.  http://p.sf.net/sfu/bobj-july
 ___
 sisuite-users mailing list
 sisuite-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/sisuite-users



--
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day 
trial. Simplify your report design, integration and deployment - and focus on 
what you do best, core application coding. Discover what's new with
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] problem installing SI 4.0.2 Ubuntu Hardy

2009-10-20 Thread Bernard Li

Hi Eric:

Maybe you can try the deb from here:

http://bison.csm.ornl.gov/repos/debian-4-x86_64/systemimager-initrd-template-amd64_4.0.2_all.deb

Traditionally the OSCAR project (at least when I was maintaining it)
always use the unstable version of SystemImager. The releases of
SystemImager has generally been solid even though it is marked
unstable -- it just means that newer features and other bleeding
edge technologies are included, but it does not necessarily mean the
imaging core is unstable.

Anyway, you may have issues using 4.0.2 with newer hardware, so when
you do, you don't really have an option but to use the newer code.

Please test the above deb and if it works fine for you, I'll upload it
to SourceForge.

Thanks,

Bernard

On Sun, Oct 18, 2009 at 10:29 PM, Eric Ross er...@anl.gov wrote:
Bernard,

Also, it is probably a bad idea to convert the
x86_64 rpm to a deb anyway.

I agree. It was more of a last resort on my end. So here's the whole
story. I checked out the stable release from svn with the intent to
build my debs from source. I ran a make deb, and unfortunately the
build would fail (I will provide more detail in my next email, I'm not
near my build box at the moment).

If you do build it, please send me a copy so that I can upload it to
SourceForge.

I'll be more than happy to -- hopefully the problem I mentioned above
will be easy to solve.

P.S. Have you considered using 4.1.6 instead? That version has the
systemimager-boot-amd64-standard debian package.

Ah, no I didn't know that. I would prefer to use the stable release
though -- but I can start my testing with this release :)

Thanks for the help!

Cheers,

-
Eric Ross
Linux Systems Administrator
Argonne National Laboratory
Computing and Information Systems
Phone: 630.252.2600
-

--
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Newer Kernel (2.6.31) and LVM

2009-10-20 Thread Bernard Li

Hey Jeremy, others:

I just checked some code into trunk, can you guys please give it a spin?

http://therealms.org/oss/systemimager/snapshots/

Changelog:
- Update i386 kernel to 2.6.31.3
- Removed 50compile-with-gcc-4-3.patch since it is no longer necessary
for 2.6.31.3
- Added patch for util-linux since kernel no longer exports asm/page.h
(patch taken from Gentoo Bug 168278)
- Update tar to 1.22
- Update kexec-tools to 2.0.1

For those curious, the latest code builds on Fedora 9 i386 w/ gcc 4.3.0.

Thanks,

Bernard

On Wed, Oct 14, 2009 at 4:13 PM, Jeremy Enos je...@ncsa.uiuc.edu wrote:
 I don't have any i386 machines anymore, now that I think about it.

 Bernard Li wrote:

 Hi Jeremy:

 On Tue, Oct 13, 2009 at 2:35 AM, Jeremy Enos je...@ncsa.uiuc.edu wrote:



 Bernard- it works!  thank you very much for fixing this!  I'm back in
 business.


 Glad to know that it works for you.  Can you (or anybody else)
 volunteer to test on the i386 platform as well?  If so then I will
 build and post them.  And if all looks good then I'll check the
 changes into trunk.

 Thanks,

 Bernard



--
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay 
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Newer Kernel (2.6.31) and LVM

2009-10-20 Thread Bernard Li

Hey Jeremy:

On Tue, Oct 20, 2009 at 5:53 PM, Jeremy Enos je...@ncsa.uiuc.edu wrote:

 I would- but not sure when i386 nodes are going to cross my path again

Actually I need testing on both x86_64 and i386 nodes.  I checked in
new code that needs testing (since I don't really have a testing
framework now).

Thanks,

Bernard

--
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay 
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Newer Kernel (2.6.31) and LVM

2009-10-14 Thread Bernard Li

Hi Matteo:

On Wed, Oct 14, 2009 at 12:12 AM, d4lamar d4la...@gmail.com wrote:

 Ok I'll do the job.
 Is there a deadline ?
 I'm quite busy.

How much time do you need?  If you could just provide the unified
version of the patch you submitted, I can do a better job at reviewing
it.  Then we'll see what else needs to be done.

Cheers,

Bernard

--
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay 
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Newer Kernel (2.6.31) and LVM

2009-10-14 Thread Bernard Li

Hi Jeremy:

On Tue, Oct 13, 2009 at 2:35 AM, Jeremy Enos je...@ncsa.uiuc.edu wrote:

 Bernard- it works!  thank you very much for fixing this!  I'm back in
 business.

Glad to know that it works for you.  Can you (or anybody else)
volunteer to test on the i386 platform as well?  If so then I will
build and post them.  And if all looks good then I'll check the
changes into trunk.

Thanks,

Bernard

--
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay 
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Newer Kernel (2.6.31) and LVM

2009-10-13 Thread Bernard Li

Hello:

On Tue, Oct 13, 2009 at 12:56 AM, d4lamar d4la...@gmail.com wrote:

 In my environment I want to be able to reinstall a system without
 losing data on certain LVM disks.
 The format=no option isn't enough to guarantee that data will not be
 losed. I doesn't know how LVM disks are created if they are contiguous
 space on a disk or something different so to preserve data I have to
 do not delete VolumeGroups and LogicalVolumes that contains my data.

 So I written the new option recreate=no for LogicalVolume Definition
 on autoinstallscript.conf.

 This is the patch to Server.pm module to make it works.

 I hope It will be useful for other users and included in the new version.

Thanks for your contribution.  Would it be possible for you to
generate a unified patch?  SImply doing `svn diff  patch` in the
working directory would generate such a patch.

Also, is your patch based on trunk or a released version?

Lastly, do you think you can provide patches to update the man pages
for autoinstallscript?  Do we also need to update the
autoinstallscript.conf template?

Cheers,

Bernard

--
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay 
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Newer Kernel (2.6.31) and LVM

2009-10-12 Thread Bernard Li

Dear all:

I just built the code from trunk with the latest Linux kernel
(2.6.31.3) for x86_64 and updated LVM to version 2.2.02.53 (replacing
device-mapper as a result).

The source tarball and RPMs (built on CentOS 5.3 x86_64) can be found here:

http://therealms.org/oss/systemimager/snapshots/

I haven't had a chance to test these so use at your own risk.  Please
let me know if this solves your problems.

I am planning to create a new unstable release 4.1.8 soon, so if
there are other minor patches that you would like to see checked in,
please let me know.

Thanks,

Bernard

On Fri, Sep 25, 2009 at 12:42 PM, Ronny Boesger [ISPpro Internet KG]
r...@isppro.de wrote:
 Hi Bernard,

 ok, i am on the way in getting 2.6.31.1 and lvm newset version in use.
 lvm did a merge of devmapper and lvm, so i changed that rul file
 accordingly.

 All went finde, i am near the end, but i got a error, i hope anybody can
 give me a hint on this, please find followed the last lines of a make all:

 If i succeed in getting this to work, i will try to update some other
 utils to a newer version, as this are somewhat really old versions in SI :)

 Thanks,
 Ronny

 ==CUT==
 I: Using ld-linux-x86-64.so.2 as dynamic linker.
 I: library reduction pass 1
 Objects: reiserfstune jfs_mkfs jfs_tune ssh scp ssh-keygen sshd mkfs.xfs
 xfs_db lvchange udevinfo torrentinfo-console mount dhclient udp-receiver
 umount rsync module_usb hotplug module_ieee1394 bittorrent-console
 module_pci udevd udevadm module_scsi ash mountpoint bc tar gzip discover
 kexec hformat mkdosfs mke2fs tune2fs parted mkswap sfdisk blockdev mdadm
 mkreiserfs
 Object: sbin/reiserfstune
 Object: sbin/jfs_mkfs
 Object: sbin/jfs_tune
 Object: usr/bin/ssh
 Object: usr/bin/scp
 Object: usr/bin/ssh-keygen
 Object: usr/sbin/sshd
 Object: sbin/mkfs.xfs
 Object: sbin/xfs_db
 Object: sbin/lvchange
 Object:
 /data1/SystemImager-BUILD/systemimager.trunk/initrd_source/build_dir/usr/bin/udevinfo
 Object:
 /data1/SystemImager-BUILD/systemimager.trunk/initrd_source/build_dir/bin/torrentinfo-console
 Object:
 /data1/SystemImager-BUILD/systemimager.trunk/initrd_source/build_dir/bin/mount
 Object:
 /data1/SystemImager-BUILD/systemimager.trunk/initrd_source/build_dir/sbin/dhclient
 Object:
 /data1/SystemImager-BUILD/systemimager.trunk/initrd_source/build_dir/bin/udp-receiver
 Object:
 /data1/SystemImager-BUILD/systemimager.trunk/initrd_source/build_dir/bin/umount
 Object:
 /data1/SystemImager-BUILD/systemimager.trunk/initrd_source/build_dir/bin/rsync
 Object:
 /data1/SystemImager-BUILD/systemimager.trunk/initrd_source/build_dir/sbin/module_usb
 Object:
 /data1/SystemImager-BUILD/systemimager.trunk/initrd_source/build_dir/sbin/hotplug
 Object:
 /data1/SystemImager-BUILD/systemimager.trunk/initrd_source/build_dir/sbin/module_ieee1394
 Object:
 /data1/SystemImager-BUILD/systemimager.trunk/initrd_source/build_dir/bin/bittorrent-console
 Object:
 /data1/SystemImager-BUILD/systemimager.trunk/initrd_source/build_dir/sbin/module_pci
 Object:
 /data1/SystemImager-BUILD/systemimager.trunk/initrd_source/build_dir/sbin/udevd
 Object:
 /data1/SystemImager-BUILD/systemimager.trunk/initrd_source/build_dir/sbin/udevadm
 Object:
 /data1/SystemImager-BUILD/systemimager.trunk/initrd_source/build_dir/sbin/module_scsi
 Object:
 /data1/SystemImager-BUILD/systemimager.trunk/initrd_source/build_dir/bin/ash
 Object:
 /data1/SystemImager-BUILD/systemimager.trunk/initrd_source/build_dir/bin/mountpoint
 Object: bin/bc
 Object: bin/tar
 Object: bin/gzip
 Object: sbin/discover
 Object: sbin/kexec
 Object: sbin/hformat
 Object: sbin/mkdosfs
 Object: sbin/mke2fs
 Object: sbin/tune2fs
 Object: sbin/parted
 Object: sbin/mkswap
 Object: sbin/sfdisk
 Object: sbin/blockdev
 Object: sbin/mdadm
 Object: sbin/mkreiserfs
 945 symbols, 936 unresolved
 Traceback (most recent call last):
  File
 /data1/SystemImager-BUILD/systemimager.trunk/initrd_source/mklibs,
 line 480, in module
    raise No library provides non-weak  + symbol
 TypeError: exceptions must be classes or instances, not str
 make: ***
 [/data1/SystemImager-BUILD/systemimager.trunk/tmp/boel_binaries.tar.gz]
 Error 1
 ==CUT==

 Bernard Li schrieb:
 Hi Ronny:

 Since you are rebuilding SystemImager from scratch anyway, I suggest
 you update both the kernel and the lvm package to their respective
 newest version.  If you need specific instructions to do that, please
 let me know.

 Don't have a lot of time to spend on SystemImager these days, but if
 you could test the configuration with newer kernel + LVM and provide
 us with a patch, I could perhaps check it into the repository.

 Thanks,

 Bernard

 On Fri, Sep 25, 2009 at 10:15 AM, Ronny Boesger [ISPpro Internet KG]
 r...@isppro.de wrote:
 Hello,

 first, nice work with systemimager :)

 I have a issue with systemimager trunk and a newer kernel. I have build
 a newer kernelk 2.6.31, copied

Re: [sisuite-users] Deployment fails on pvcreate

2009-10-02 Thread Bernard Li

Maybe it has to do with lvm:

http://www.mail-archive.com/sisuite-users@lists.sourceforge.net/msg05064.html

Cheers,

Bernard

On Fri, Oct 2, 2009 at 2:51 PM, Jeremy Enos je...@ncsa.uiuc.edu wrote:
 Hi Tony-
 I'm having precisely the same problem as you had here.  Did you ever
 find a solution?  It's like there's a bug w/ pvcreate or something.

 (I've tried 4.1.6, UYOK- no help)

    Jeremy

 Tony Spencer wrote:
 My progress in deploying an image to a new server has moved on.
 The autoinstall script can now see my /dev/sda partition.
 It seems to partition the disk fine but then fails on the pvcreate part.
 It complains it can't see /dev/sda2, but when I inspect the disk with
 parted or fdisk I  can see it.

 Below is the part of the logging where it errors out.

 Thanks
 Tony


 get_arch
 Partitioning /dev/sda...
 Old partition table for /dev/sda:
 Error: Unable to open /dev/sda - unrecognised disk label.
 dd if=/dev/zero of=/dev/sda bs=512 count=1 || shellout
 1+0 records in
 1+0 records out
 blockdev --rereadpt /dev/sda
 parted -s -- /dev/sda mklabel msdos || shellout
 Creating partition /dev/sda1.
 parted -s -- /dev/sda mkpart primary 0 196 || shellout
 parted -s -- /dev/sda set 1 boot on
 Creating partition /dev/sda2.
 parted -s -- /dev/sda mkpart primary 196 71999 || shellout
 parted -s -- /dev/sda set 2 lvm on
 New partition table for /dev/sda:
 parted -s -- /dev/sda print
 Disk geometry for /dev/sda: 0kB - 72GB
 Disk label type: msdos
 Number Start End Size Type File system Flags
 1 1kB 196MB 196MB primary boot
 2 196MB 72GB 72GB primary lvm
 Load software RAID modules.
 modprobe: module raid5 not found.
 modprobe: failed to load module raid5
 modprobe: module raid6 not found.
 modprobe: failed to load module raid6
 Load device mapper driver (for LVM).
 insmod: cannot insert
 `/lib/modules/2.6.25-14.fc9.i686/kernel/drivers/md/dm-mod.ko': File
 exists (-1): File exists
 modprobe: failed to load module dm-mod
 Initializing partition /dev/sda2 for use by LVM.
 pvcreate -M2 -ff -y /dev/sda2 || shellout
 Device /dev/sda2 not found (or ignored by filtering).
 Killing off running processes.



 -
 This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
 Build the coolest Linux based applications with Moblin SDK  win great prizes
 Grand prize is a trip for two to an Open Source event anywhere in the world
 http://moblin-contest.org/redirect.php?banner_id=100url=/
 ___
 sisuite-users mailing list
 sisuite-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/sisuite-users



 --
 Come build with us! The BlackBerryreg; Developer Conference in SF, CA
 is the only developer event you need to attend this year. Jumpstart your
 developing skills, take BlackBerry mobile applications to market and stay
 ahead of the curve. Join us from November 9#45;12, 2009. Register now#33;
 http://p.sf.net/sfu/devconf
 ___
 sisuite-users mailing list
 sisuite-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/sisuite-users


--
Come build with us! The BlackBerryreg; Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay 
ahead of the curve. Join us from November 9#45;12, 2009. Register now#33;
http://p.sf.net/sfu/devconf
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] autoinstall failed

2009-10-02 Thread Bernard Li

Hi Joe:

On Fri, Oct 2, 2009 at 6:53 AM, Joe Koenig j...@creativeanvil.com wrote:

 Yeah, that was the first thing I tried. When renaming the file didn't
 work, I threw the -S option in, which gave the same result. If this file
 was corrupted somehow, would I have gotten as far in the process as I
 did, or would it just have not run at all? And, yes, I'm working with a
 copy of the file, just in case.

Okay this is pretty strange.

So can you take a look at the initrd.img from
/usr/share/systemimager/boot/i386/standard/ (replace i386 with x86_64
if your server is x86_64) and compare that with the initrd.img that
you were working on.  Are they the same?

Oh and one more thing.  Can you post the output of:

rpm -qa | grep systemimager

Thanks,

Bernard

--
Come build with us! The BlackBerryreg; Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay 
ahead of the curve. Join us from November 9#45;12, 2009. Register now#33;
http://p.sf.net/sfu/devconf
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] autoinstall failed

2009-10-01 Thread Bernard Li

Hi Joe:

On Thu, Oct 1, 2009 at 11:29 AM, Joe Koenig j...@creativeanvil.com wrote:

 Thanks for the response. I am using systemimager 4.0.2 on CentOS 5.3, 64-bit
 on the server. The client was RHEL 4.5 64-bit. The golden client is 64-bit
 -- the same machine that I'm trying to install back onto now. Kernel is
 BOEL.

Try again with UYOK and see if that helps.

BTW, you might want to give SystemImager 4.1.6 a shot -- the newer
code might help.

Cheers,

Bernard

--
Come build with us! The BlackBerryreg; Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay 
ahead of the curve. Join us from November 9#45;12, 2009. Register now#33;
http://p.sf.net/sfu/devconf
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] autoinstall failed

2009-10-01 Thread Bernard Li

Just rename the file to .gz.

Regards,

Bernard

P.S. You probably want to make a backup of the original initrd.img
before you play with it.

On Thu, Oct 1, 2009 at 12:47 PM, Joe Koenig j...@creativeanvil.com wrote:
 On 10/1/09 2:19 PM, Bernard Li wrote:

 Hi Joe:

 On Thu, Oct 1, 2009 at 12:00 PM, Joe Koenig j...@creativeanvil.com wrote:



 Thanks for the suggestion. I'm in a bit of a bind though because the golden
 client doesn't exist anymore. Am I just out of luck on being able to restore
 that image without the golden client?


 Well, then the only way around it is to hack your initrd.img.
 Basically it is a gzip compressed file which after decompression can
 be mounted via the loopback.  Then you can peek inside to see whether
 libuuid.so.1 is indeed present.

 BTW, the environment when you drop out to shell and the environment of
 initrd are not necessarily the same -- so it's possible that the
 library is not present in the initrd but present on the filesystem
 after BOEL binaries were retrieved.

 Cheers,

 Bernard

 Thanks for the tip. I made a copy of the file and did the following:

 [r...@file initrd]# gunzip -S .img initrd.img

 gunzip: initrd.img: not in gzip format

 Any ideas? Could this be the issue? A corrupt initrd.img perhaps?

 Thanks again for all your help.

 --
 Come build with us! The BlackBerryreg; Developer Conference in SF, CA
 is the only developer event you need to attend this year. Jumpstart your
 developing skills, take BlackBerry mobile applications to market and stay
 ahead of the curve. Join us from November 9#45;12, 2009. Register now#33;
 http://p.sf.net/sfu/devconf
 ___
 sisuite-users mailing list
 sisuite-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/sisuite-users



--
Come build with us! The BlackBerryreg; Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay 
ahead of the curve. Join us from November 9#45;12, 2009. Register now#33;
http://p.sf.net/sfu/devconf
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Problem imaging Dell PowerEdge R410

2009-09-29 Thread Bernard Li

Hi Dmitri:

On Tue, Sep 29, 2009 at 12:16 PM, Gerasimatos, Dimitrios V (343K)
dimitrios.v.gerasima...@jpl.nasa.gov wrote:

 I have a RHEL4U8 image that works for imaging Dell 340s to Dell T3400s.
 This same image gets the Dell PowerEdge R410 up to the point where it
 correctly gets its network configuration information from the System Imager
 Server and then fails with the following error:
 ...
 Read_kernel_append_parameters
 Start_network
 ifconfig: SIOCSIFADDR: No such Device
 I couldn't configure the network using your pre-boot settings
 Device: eth0
 IPAddr: (obtained correct IP from SIS)
 Netmask: (obtained correct netmask from SIS)
 Broadcast: (obtained correct broadcast from SIS)
 Killing off running processes.
 ...
 Installation failed!!  Stopping report task
 nc: cannot connect to remote host (IP of server): Network is unreachable


 The R410 has a Broadcom NetXtreme II BCM5716 Gigabit Ethernet card.

How about the Dell 340s to Dell T3400s servers?  What nics do they
have?  And more importantly, what kernel modules do they use?

Cheers,

Bernard

--
Come build with us! The BlackBerryreg; Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay 
ahead of the curve. Join us from November 9#45;12, 2009. Register now#33;
http://p.sf.net/sfu/devconf
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Newer Kernel (2.6.31) and LVM

2009-09-25 Thread Bernard Li

Hi Ronny:

Since you are rebuilding SystemImager from scratch anyway, I suggest
you update both the kernel and the lvm package to their respective
newest version.  If you need specific instructions to do that, please
let me know.

Don't have a lot of time to spend on SystemImager these days, but if
you could test the configuration with newer kernel + LVM and provide
us with a patch, I could perhaps check it into the repository.

Thanks,

Bernard

On Fri, Sep 25, 2009 at 10:15 AM, Ronny Boesger [ISPpro Internet KG]
r...@isppro.de wrote:
 Hello,

 first, nice work with systemimager :)

 I have a issue with systemimager trunk and a newer kernel. I have build
 a newer kernelk 2.6.31, copied the modules to boel_binaries.tar.gz file
 and the updated the kernel.

 All is working perfectly, but lvm makes problems. The partition is
 created through parted correctly, but then the pvcreate command fails:

 # pvcreate -M2 -ff -y /dev/sda2
  Device /dev/sda2 not found (or ignored by filtering).

 This works with a older kernel (2.6.24) without any problem, so it must
 be kernel related.

 I read and found, that the sysfs may be changed, as in a repository of
 lvm i found a change short after the release of the plain old lvm in
 systemimager :)
 The patch is attached as diff file,
 The url for the original file is:
 http://sources.redhat.com/ml/lvm2-cvs/2007-10/msg8.html

 I first ask, if anyone have a clue, if this is correct, i don`t want to
 use UYOK, as 1 kernel for all distributions is simpler to manage in sum.

 I need a new kernel, as we use new hardware /intel dg43gt mainboard, and
 some others with ICH10R).


 BTW: Are there plans, to upgrade systemimager to a newer kernel, lvm
 etc. are here testers or so needed ?

 thanks,
 ronny


 
 # Added by Ronny Boesger, this is for supporting newer kernels.
 # in my case, its 2.6.31
 

 *** lib/filters/filter-sysfs.c  2007-04-26 18:44:58.0 +0200
 --- lib/filters/filter-sysfs.c.new      2009-09-25 18:48:07.0 +0200
 ***
 *** 20,31 

  #include dirent.h

 ! static int _locate_sysfs_blocks(const char *proc, char *path, size_t len)
  {
        char proc_mounts[PATH_MAX];
 -       int r = 0;
        FILE *fp;
        char *split[4], buffer[PATH_MAX + 16];

        if (!*proc) {
                log_verbose(No proc filesystem found: skipping sysfs filter);
 --- 20,33 

  #include dirent.h

 ! static int _locate_sysfs_blocks(const char *proc, char *path, size_t len,
 !                               unsigned *sysfs_depth)
  {
        char proc_mounts[PATH_MAX];
        FILE *fp;
        char *split[4], buffer[PATH_MAX + 16];
 +       const char *sys_mnt = NULL;
 +       struct stat info;

        if (!*proc) {
                log_verbose(No proc filesystem found: skipping sysfs filter);
 ***
 *** 46,55 
        while (fgets(buffer, sizeof(buffer), fp)) {
                if (dm_split_words(buffer, 4, 0, split) == 4 
                    !strcmp(split[2], sysfs)) {
 !                       if (dm_snprintf(path, len, %s/%s, split[1],
 !                                        block) = 0) {
 !                               r = 1;
 !                       }
                        break;
                }
        }
 --- 48,54 
        while (fgets(buffer, sizeof(buffer), fp)) {
                if (dm_split_words(buffer, 4, 0, split) == 4 
                    !strcmp(split[2], sysfs)) {
 !                       sys_mnt = split[1];
                        break;
                }
        }
 ***
 *** 57,63 
        if (fclose(fp))
                log_sys_error(fclose, proc_mounts);

 !       return r;
  }

  /*
 --- 56,125 
        if (fclose(fp))
                log_sys_error(fclose, proc_mounts);

 !       if (!sys_mnt) {
 !       log_error(Failed to find sysfs mount point);
 !       return 0;
 !       }
 !
 !       /*
 !        * unified classification directory for all kernel subsystems
 !        *
 !        * /sys/subsystem/block/devices
 !        * |-- sda - 
 ../../../devices/pci:00/:00:1f.2/host0/target0:0:0/0:0:0:0/block/sda
 !        * |-- sda1 - 
 ../../../devices/pci:00/:00:1f.2/host0/target0:0:0/0:0:0:0/block/sda/sda1
 !        *  `-- sr0 - 
 ../../../devices/pci:00/:00:1f.2/host1/target1:0:0/1:0:0:0/block/sr0
 !        *
 !        */
 !       if (dm_snprintf(path, len, %s/%s, sys_mnt,
 !       subsystem/block/devices) = 0) {
 !       if (!stat(path, info)) {
 !       *sysfs_depth = 0;
 !       return 1;
 !       }
 !         }
 !
 !       /*
 !        * block subsystem as a class
 !        *
 !        * /sys/class/block
 !        * |-- sda - 
 ../../devices/pci:00/:00:1f.2/host0/target0:0:0/0:0:0:0/block/sda
 !        * |-- sda1 -

Re: [sisuite-users] Problem loading my image after pxelinux.cgf/default load

2009-09-10 Thread Bernard Li

Hi Alex:

2009/9/10 Alexandre González a...@rianxosencabos.com:

 I can't copypaste it, but basically it try to load pxelinux.cfg/my_hd_id,
 pxelinux.cfg/another_id, pxelinux.cg/... over 8 times, and then try to load
 pxelinux.cfg/default it can't do it and load my default SO on the computer.

What do you mean by it can't do it?  Are you using SystemImager's
pxelinux.cfg/default file?  Can you post the contents of that file?

Cheers,

Bernard

--
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day 
trial. Simplify your report design, integration and deployment - and focus on 
what you do best, core application coding. Discover what's new with 
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Problem loading my image after pxelinux.cgf/default load

2009-09-09 Thread Bernard Li

Hi Alex:

2009/9/7 Alexandre González a...@rianxosencabos.com:

 My problem is that I have a server with dhcp3, atftp and, of course
 systemimager, but when I try to load my gold client image from another
 client I can't... the pxelinux.cfg/default doesn't do anything, so I use
 PROMPT 1 on it, and now I can write after the boot:  string. So I used it
 to write the full path (relative path at my system, but full in the tftp)
 something as i386/mdealer.img/kernel and it boots, but It give me a kernel
 panic. Same problem booting i386/standard/kernel.

Can you show us what's on your screen when your client tries to
pxeboot?  Did it just sit at the prompt without loading the default
boot process?

Cheers,

Bernard

--
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day 
trial. Simplify your report design, integration and deployment - and focus on 
what you do best, core application coding. Discover what's new with 
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Hi

2009-08-17 Thread Bernard Li

Hi Joe:

On Mon, Aug 17, 2009 at 6:19 AM, Varghese, Anjovarghese...@pg.com wrote:

 How to respond to the reply I got for a query I submitted?

It looks like you are on digest mode for the mailing-list.  I would
suggest using normal mode so that you get emails immediately.  It is
not like the mailing-list is exactly high traffic.

Cheers,

Bernard

--
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day 
trial. Simplify your report design, integration and deployment - and focus on 
what you do best, core application coding. Discover what's new with 
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] SATA Raid0 Drivers

2009-08-14 Thread Bernard Li

Hi Joe:

On Fri, Aug 14, 2009 at 6:59 AM, Varghese, Anjovarghese...@pg.com wrote:

 I have a Raid0 Sata Configuration for the Oscar 5.0 implementation on RHEL
 4.6 x86_64, But I have problems in booting up compute nodes, if I use the
 UYOK option it will do the imaging but after booting I get a kernel panic
 message and the compute nodes hangs, if I don’t use YUOK it won’t recognize
 the disks. It would be of great help if someone can guide me on a workaround
 for this. I’m attaching the output of lsmod, lspci and /proc/pci to identify
 the hardware, I’m using the LSI Logic SAS card. I haven’t load any drivers
 manually, when I installed the head node, all drivers are picked up
 automatically.

Have you tried this?

http://www.mail-archive.com/oscar-us...@lists.sourceforge.net/msg07280.html

Also, how did you setup RAID0?  AFAIK OSCAR does not support setting
up RAID for compute nodes (even though SystemImager supports it).

Cheers,

Bernard

--
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day 
trial. Simplify your report design, integration and deployment - and focus on 
what you do best, core application coding. Discover what's new with 
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] SystemImager + LTSP

2009-08-14 Thread Bernard Li

Hi Àlex:

This looks interesting.

Have you done any performance comparisons between SystemImager+LTSP vs
standard SystemImager (like time it takes to image node and potential
load on NFS server when imaging multiple systems).

SystemImager has been traditionally used to deploy images to a large
number of systems at the same time, so performance and scalability is
quite an important metric for the software.

Cheers,

Bernard

On Fri, Aug 14, 2009 at 1:50 AM, Àlex Magaz Graçaal...@asic.udl.cat wrote:
 Hi,

 Some months ago I was working on integrating SystemImager into LTSP. The
 idea was to run the installation scripts from an LTSP environment instead of
 running them from a custom initrd.

 Why do this? Well, right now, if you want to modify some of this scripts,
 you have to regenerate the initrd and put it into the right place. As you
 usually make mistakes you have to repeat the process several times, which is
 a bit tedious. Things become worse when you need to regenerate the BOEL
 binaries tar ball, for example to update some of its utilities. It also a
 pain when something fails in the middle of an installation and you have to
 find out what happened from a so limited busy box environment.

 So, what do we get from this? With LTSP we get a chroot of a standard
 distribution mounted through NFS, which means we can modify any file and
 install/upgrade any package in a really easy way. This makes it easier to
 test new functionality and debug problems.

 With the attached patch I've succeeded doing installations with both rsync
 and BitTorrent, but there's still a lot of work do. I've not tested
 multicast and SSH installations, support for all installation parameters is
 incomplete and some other things don't work as they should.

 As I don't think I'll be able to work on this anymore, I drop it here in the
 hope someone finds it useful. Apart from the patch, I've also attached the
 instructions on how to integrate the SystemImager scripts into the LTSP
 chroot.

 Cheers,
 Àlex


 --
 Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
 trial. Simplify your report design, integration and deployment - and focus
 on
 what you do best, core application coding. Discover what's new with
 Crystal Reports now.  http://p.sf.net/sfu/bobj-july
 ___
 sisuite-users mailing list
 sisuite-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/sisuite-users



--
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day 
trial. Simplify your report design, integration and deployment - and focus on 
what you do best, core application coding. Discover what's new with 
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Error opening /dev/sda: No medium found

2009-08-13 Thread Bernard Li

Hi Richard:

Have you tried UYOK?

Cheers,

Bernard

On Thu, Aug 13, 2009 at 11:49 AM, RICHARD GREENgree...@u.washington.edu wrote:
 Hi everyone, I have successfully created a golden client and retrieved it to 
 my systemimager (3.7.5-1) server, I have done this in the past with few 
 problems. This new hardware that I am placing the image on is the exact same 
 hardware that I got the image off of.  When I boot a node to load the image, 
 it seems to get stuck in not identifying the scsi drive (see error message 
 below)

 run_autoinstall_script
          /scripts/oscarnode11.sh

         get_arch


         Partitioning /dev/sda...
         Old partition table for /dev/sda:
         Error: Error opening /dev/sda: No medium found
         dd if=/dev/zero of=/dev/sda bs=512 count=1 || shellout
         dd: /dev/sda: No medium found
         Killing off running processes.

 There is definitely an /dev/sda drive, can anyone suggest what I might be 
 missing? Any help or suggestions is muchly appreciated Thanks -Rich


 [r...@localhost ~]# fdisk -l

 Disk /dev/sda: 146.8 GB, 146815733760 bytes
 255 heads, 63 sectors/track, 17849 cylinders
 Units = cylinders of 16065 * 512 = 8225280 bytes

    Device Boot      Start         End      Blocks   Id  System
 /dev/sda1   *           1          13      104391   83  Linux
 /dev/sda2              14       17849   143267670   8e  Linux LVM






 --
 Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
 trial. Simplify your report design, integration and deployment - and focus on
 what you do best, core application coding. Discover what's new with
 Crystal Reports now.  http://p.sf.net/sfu/bobj-july
 ___
 sisuite-users mailing list
 sisuite-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/sisuite-users


--
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day 
trial. Simplify your report design, integration and deployment - and focus on 
what you do best, core application coding. Discover what's new with 
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] systemimager

2009-07-29 Thread Bernard Li

Hi Stefan:

On Wed, Jul 29, 2009 at 5:53 AM, Stefan
Obermeierstefan.oberme...@biozentrum.uni-wuerzburg.de wrote:

 I just read this thread
 http://www.mail-archive.com/sisuite-users@lists.sourceforge.net/msg04858.html
 and wondered how to run this patch against the UseYourOwnKernel.pm?
 I have the same problem with sles11.

First, download the patch, then cd into
/usr/lib/systemimager/perl/SystemImager/ and run:

patch -p0  UseYourOwnKernel.pm.patch

It would be great if you can confirm that this works so that I can
check this into the code repository.

P.S. Please reply back to the mailing-list -- you may need to
subscribe to it if you are not already a member.

Cheers,

Bernard

--
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day 
trial. Simplify your report design, integration and deployment - and focus on 
what you do best, core application coding. Discover what's new with 
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Ubuntu 9.04 - dm-mod - LVM2 - SI 4.1.6

2009-07-21 Thread Bernard Li

Hi Phil:

On Tue, Jul 21, 2009 at 10:19 AM, Phil Forrestphilforr...@gmail.com wrote:

 I'm attempting to image Ubuntu 9.04 64-bit clients via PXE and UYOK (and a
 Ubuntu 9.04 32-bit image server, but I don't think that matters). PXE boot
 is working wonderfully but it appears the imaging master script attempts to
 modprobe a non-existent dm-mod (Device Mapper) kernel module.

Have you tried imaging the node with the BOEL kernel instead of the
UYOK kernel?  It is a bit old but it might just do the trick.

Cheers,

Bernard

--
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Saving image to USB Drive?

2009-07-09 Thread Bernard Li

Hi Jim:

On Thu, Jul 9, 2009 at 8:43 AM, Jim O'Beirnejimobei...@gmail.com wrote:

 How would I save the golden_image to a dir on an external hard drive? I am
 trying to create an image of a system that has a large hard drive to a
 server with a smaller hard drive BUT has large external HDD. Have search
 docs for help to no avail. Or, could I save the image from the client to the
 ext HDD on the client?

Okay, let me try to rephrase your question -- basically you are trying
to take a golden image of a client with a larger HD than that of the
image server's HD.  The good news is that on your image server, you
have an external HD that has enough capacity to hold the client image.

If that's the case, then all you need to do is get SystemImager to
save the image to the external HD.

SystemImager stores all its images to the directory
/var/lib/systemimager/images, so perhaps what you can do is move the
contents of that directory to your external HD and then mount the
external HD as /var/lib/systemimager/images.  From then on any images
stored in SystemImager will be stored on the external HD instead of
the internal HD of your image server.

However, that means the external HD need to be pretty much connected
to the server all the time for SystemImager to work at all, so take
heed of that.  Also, since the images are stored on the USB drive, the
speed of imaging will take a substantial hit.

Good luck,

Bernard

--
Enter the BlackBerry Developer Challenge  
This is your chance to win up to $100,000 in prizes! For a limited time, 
vendors submitting new applications to BlackBerry App World(TM) will have
the opportunity to enter the BlackBerry Developer Challenge. See full prize  
details at: http://p.sf.net/sfu/Challenge
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] logging of post-install scripts

2009-07-08 Thread Bernard Li

Hi Johan:

On Wed, Jul 8, 2009 at 6:32 AM, Johan Huysmansjohan.huysm...@inuits.be wrote:

 I'm currently using the 4.0.2 systemimager on CentOS5.

 I've created several post-install scripts but there is no logging of the
 messages and errors in these post-install scripts.
 The messages appear on the screen, but they do not appear in the logging
 through syslog, which I'm using.


 Is there a way so I can see which messages are printed and which errors
 occur in the post-install scripts?
 This should be very helpfull.

Not sure about syslog, but if you setup logging via si_monitor, the
messages should show up:

http://wiki.systemimager.org/index.php/Monitoring

P.S. Not sure why but your post appeared to have been posted twice.

Cheers,

Bernard

--
Enter the BlackBerry Developer Challenge  
This is your chance to win up to $100,000 in prizes! For a limited time, 
vendors submitting new applications to BlackBerry App World(TM) will have
the opportunity to enter the BlackBerry Developer Challenge. See full prize  
details at: http://p.sf.net/sfu/Challenge
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] No luck with Fedora 11

2009-06-29 Thread Bernard Li

Hi Luc:

On Mon, Jun 29, 2009 at 12:55 PM, Luc Lalondeluc.lalo...@polymtl.ca wrote:

 However, for the life of me, I can't seem to understand what is wrong with
 the one that was generated by the Fedora 11 golden client.  Here's a copy if
 you can take a quick look...

I think you forgot to attach the file?

Cheers,

Bernard

--
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] No luck with Fedora 11

2009-06-26 Thread Bernard Li

Hi Luc:

On Thu, Jun 25, 2009 at 8:52 AM, Luc Lalondeluc.lalo...@polymtl.ca wrote:

 I'm not able to get SystemImager 4.0.2 working with Fedora 11.

 When I boot the client for re-installation with Bittorrent I get a df
 error about missing /a/boot...

That's not really enough information to go on -- can you post the
install log to pastebin and reference it in a reply?

You could get the install log by setting up monitoring on the image
server -- for more information look up si_monitor.

Also, have you tried using the rsync transport and see if that works?

Cheers,

Bernard

--
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] tftp timeout during pxe booting

2009-05-07 Thread Bernard Li

Hi Joe:

On Thu, May 7, 2009 at 9:08 AM, Greenseid, Joseph M (IS)
joseph.greens...@ngc.com wrote:

 i am using SIS as part of the OSCAR cluster management software.  i am
 running into a problem where when i network boot a node to install it, it
 gets an address from dhcp just fine, then tftp starts up, it finds the
 default config file, but then when the server tries to send over the kernel,
 it hangs.  on the client side, it's just stopped at the boot prompt.  on
 the server, /var/log/messages is filled with:

 atftpd[20082]: serving kernel to 192.168.1.104:57100
 atftpd[20082]: timeout: retrying...
 atftpd[20082]: serving kernel to 192.168.1.104:57100
 atftpd[20082]: timeout: retrying...

 on and on.

 i've checked all the standard stuff -- no firewalls, the kernel file is
 world readable, i even connected the two systems with a cross-over cable in
 case there was something goofy on the switch.

How about SELinux?  Has it been disabled?

 i've used this same version of OSCAR to install a handful of other clusters,
 and have never had any trouble.  i don't know what's up.  has anyone seen
 anything like this or have any suggestions of where i might look to see
 what's up?

Don't have any specific ideas, but you can test to see if your tftp
server is working correctly by installing a tftp client on a computer
on the same network and try troubleshooting from there.

Cheers,

Bernard

--
The NEW KODAK i700 Series Scanners deliver under ANY circumstances! Your
production scanning environment may not be a perfect world - but thanks to
Kodak, there's a perfect scanner to get the job done! With the NEW KODAK i700
Series Scanner you'll get full speed at 300 dpi even with all image 
processing features enabled. http://p.sf.net/sfu/kodak-com
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] userland util

2009-01-07 Thread Bernard Li

Hi Ted:

On Wed, Jan 7, 2009 at 11:21 AM, Ted Arden t...@tux.org wrote:

 after imaging, i'd like to flop the CMOS settings
 to revert back to boot off HDD or even be able to
 login to an imaged machine and make these CMOS
 mods??

 is there a utility to do this?? it's Phoenix
 BIOS.  i remember doing this back in my VA Linux
 days, but didn't know if there was another utility
 out there to do so.  i am using SLES9.

Assuming you are netbooting your machines, you can leave them to boot
off the network and control whether they re-image or boot off the
local HD via pxelinux. si_netbootmond could facilitate this.  For more
information you can look at the SystemImager service
systemimager-server-netbootmond (which controls the starting and
stopping of si_netbootmond) and the NET_BOOT_DEFAULT option in
/etc/systemimager/systemimager.conf.

Cheers,

Bernard

--
Check out the new SourceForge.net Marketplace.
It is the best place to buy or sell services for
just about anything Open Source.
http://p.sf.net/sfu/Xq1LFB
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] systemimager

2009-01-05 Thread Bernard Li

Hi Larry:

On Wed, Dec 31, 2008 at 7:34 PM, Larry Liang llian...@yahoo.ca wrote:

 I need a help to figure out the problem when I tried to si_prepareclient
 --server serverip  from  the golden client (opensuse 11.0), and I am
 getting this error:

 Evaluating initrd size to be added in the kernel boot options
 (e.g. /etc/systemimager/pxelinux.cfg/syslinux.cfg):
   suggested value - ramdisk_size=95636

 I couldn't identify your kernel file.  Please try to use --kernel option.

 I found out the help from this group to download a patch
 UseYourOwnKernel.pm.patch from

 http://therealms.org/oss/systemimager/UseYourOwnKernel.pm.patch

 When I tried to apply the patch

 # patch UseYourOwnKernel.pm.patch

 and getting this error:

 patching file UseYourOwnKernel.pm
 Hunk #1 FAILED at 460.
 Hunk #2 FAILED at 469.

 I am a newbie and this is my first time to use this tool, and wonder where I
 did wrong to apply the patch.

 Here is the example I refer to:

 http://www.mail-archive.com/sisuite-users@lists.sourceforge.net/msg04858.html

Which version of SystemImager are you running?  The patch assumes you
are using SystemImager 4.1.6.

Cheers,

Bernard

--
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Verbose image installation

2008-12-18 Thread Bernard Li

Hi Brett:

On Thu, Dec 18, 2008 at 9:57 AM, Brett K. Chain brettkch...@yahoo.com wrote:

 I have a seemingly simple question regarding a verbose image installation. 
 Typically when an image is being installed, it shows Quietly installing 
 image. Is there a way to make this verbose where it shows what is actively 
 being rsync'd to the client? I would like to be able to tell from the 
 client-side that the process is running.

In your autoinstallscript (found in
/var/lib/systemimager/scripts/imagename.master), comment out the
line which reads:

NO_LISTING=yes

FYI, this is controlled by the '--listing' option for
si_mkautoinstallscript.  For more information please refer to the
manpage.

Cheers,

Bernard

--
SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada.
The future of the web can't happen without you.  Join us at MIX09 to help
pave the way to the Next Web now. Learn more and register at
http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] sisuite-users Digest, Vol 31, Issue 4

Hi Adam:

On Sun, Dec 14, 2008 at 7:13 PM, Adam LaMee adamla...@gmail.com wrote:

 Bernard, thanks for the reply.  I just reinstalled the server and new
 systemimager 4.0.2 installed (the error message has been corrected, thanks
 again).  si_mkbootserver is still stopping at the same point: in.tftpd not
 found.  Ports are open on the client and I am able to do a standard network

Do you have 'tftp' in your path?  If not, you need to install that.

Cheers,

Bernard

--
SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada.
The future of the web can't happen without you.  Join us at MIX09 to help
pave the way to the Next Web now. Learn more and register at
http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] suse 11 + I couldn't identify your kernel file

Hi Joydeep:

On Fri, Dec 12, 2008 at 2:01 AM, J. Bakshi joyd...@infoservices.in wrote:

 I am totally newbie in systemimager. I am supervisibg a remote server
 running suse 11. To see the potential of system imager I hve installed
 the client packages in that server namely

 systemconfigurator-2.2.11-1
 systemimager-common-4.1.6-1
 systemimager-client-4.1.6-1

 To prepare the image I have fiven the command as root

 # si_prepareclient --server  suse11

 but after copying  some modules  it produced an error like

 ~~~
 I couldn't identify your kernel file.  Please try to use --kernel option
 `

 I have then applied the --Kernel option like

 #  si_prepareclient --server  suse11 --kernel  vmlinuz-2.6.25.11-0.1-default

 still it  reproduced the same error.

 Could  any one kindly show me the way to fix it ?

Try this patch against
/usr/lib/systemimager/perl/SystemImager/UseYourOwnKernel.pm and see if
this works (please make a backup copy of that file before you apply
the patch):

Index: UseYourOwnKernel.pm
===
--- UseYourOwnKernel.pm (revision 4466)
+++ UseYourOwnKernel.pm (working copy)
@@ -460,6 +460,7 @@
 #   2.4.19-mantis-2002.11.20 (r...@mantis) #6 Tue
Nov 19 15:15:43 CST 2002
 #   2.6.7-1-686 (dilin...@toaster.hq.voxel.net)
#1 Thu Jul 8 05:36:53 EDT 2004
 #   2.6.22.5-31-default (ge...@buildhost) #1 SMP
2007/09/21 22:29:00 UTC
+#   2.6.25.18-0.2-default #1 SMP 2008-10-21
16:30:26 +0200 [openSUSE 11.0]
 #
 my $regex =
 #   | kernel version + build machine
@@ -468,11 +469,11 @@
 #
 #   | build date
 #   `---
-'(\w{3} \w{3} \d{1,2}|\d{4}\/\d{2}\/\d{2}) '.
+'(\w{3} \w{3} \d{1,2}|\d{4}[\/-]\d{2}[\/-]\d{2}) '.
 #
 #   | build time
 #   `---
-'\d{2}:\d{2}:\d{2} \w{3,4}( \d{4})?';
+'\d{2}:\d{2}:\d{2} [\w{3,4}( \d{4})?|+\d{4}]';
 while(IN) {
# extract the `uname -r` string from the kernel file
if(m/$regex/o) {

Patch is also available here in case the inline patch does not work:

http://therealms.org/oss/systemimager/UseYourOwnKernel.pm.patch

I have tested the code with openSUSE 11.0 and that seems to work fine.

Cheers,

Bernard

--
SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada.
The future of the web can't happen without you.  Join us at MIX09 to help
pave the way to the Next Web now. Learn more and register at
http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] SystemImager and creating a VMware VM

Hi Dimitri:

On Fri, Dec 12, 2008 at 12:46 PM, Dimitri Yioulos
dyiou...@firstbhph.com wrote:

 I'm new to SystemImager. Installation on my CentOS 5.2 image server and CentOS
 golden client went fine.  Creation of the golden client image went without
 error, as did the creation of a boot CD.  Id like to create a VMware Server
 virtual machine from the image I created.  After creating the VM (trying
 Guest Operating System type Other Linux 2.6.x kerneland Other Linux), when
 I try to install from the boot CD, I get the following error:

 ISOLINUX 3.72 2008-09-25 ETCD Copyright (C) 1994-2008 H. Peter Anvin
 Unknown keyword in configuration file: f
 Unknown keyword in configuration file: BOOT.CAT:1
 Could not find kernel image: linux
 boot:

 Can I create a VMware VM from SystemImager?  Via a SystemImager CD?  Do I have
 to create the CD in a special way?

Have you tried booting the CD outside of VMware (i.e. from bare metal)
and see if it gives you the same error?

Cheers,

Bernard

--
SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada.
The future of the web can't happen without you.  Join us at MIX09 to help
pave the way to the Next Web now. Learn more and register at
http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Problem with partitioning for dual boot

On Sun, Nov 30, 2008 at 10:53 AM, Ali Nazemian alinazem...@gmail.com wrote:

 I want to build hybrid cluster network , using OSCAR and Windows hpc pack
 2008 , as u know , oscar uses system imager for deploying images to compute
 nodes, how can i change scsi.disk for changing default partiotioning
 method as i want to deploying images on hard drives that have windows hpc
 server 2008 on first drive?!

 /dev/sda1 24 ext3 /boot defaults
 /dev/sda5 128 swap
 /dev/sda6 * ext3 / defaults
 nfs_oscar:/home - nfs /home rw

Have you taken a look at this document?

http://svn.oscar.openclustergroup.org/trac/oscar/wiki/TipWindowsAndLinux

Cheers,

Bernard

--
SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada.
The future of the web can't happen without you.  Join us at MIX09 to help
pave the way to the Next Web now. Learn more and register at
http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Booting from a CD written using si_mkautoinstallcd

2008-12-12 Thread Bernard Li

Hi Mark:

On Fri, Dec 12, 2008 at 12:18 AM, Mark Hewitt mhew...@cs.york.ac.uk wrote:

 Which flavor would you like to use? [standard]: standard
 /usr/share/systemimager/boot/x86_64/standard/kernel does not exist.

Install this and try again:

http://downloads.sourceforge.net/systemimager/systemimager-x86_64boot-standard-4.0.2-1.noarch.rpm?modtime=1197135070big_mirror=1

Cheers,

Bernard

--
SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada.
The future of the web can't happen without you.  Join us at MIX09 to help
pave the way to the Next Web now. Learn more and register at
http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Booting from a CD written using si_mkautoinstallcd

Hi Mark:

On Thu, Dec 11, 2008 at 6:38 AM, Mark Hewitt mhew...@cs.york.ac.uk wrote:

 Basically when I create a .iso from an image and attempt to run it on a
 machine I get the error Kernel panic - not syncing: Oops and then the
 boot stops.

 Does anyone have any ideas as to what is going wrong here, we really
 need to get this working!

What version of SystemImager are you using?

Instead of making a bootable ISO, have you tried netbooting the client
instead?  Even if your client does not support PXE, you could use
something like etherboot.

 [r...@cortex1 cortex1-image]# si_mkautoinstallcd --out-file cortex1-1.iso
 Here is a list of available flavors:

   cortex1-image
   standard

 Which flavor would you like to use? [standard]: cortex1-image

Have you tried building an ISO with the standard flavour?

 Creating temporary working directory...
 umount: /tmp/systemimager.autoinstallcd.temp.dir/mnt: not found

This seems kind of bad...

 Pid: 635, comm: default.hotplug Not tainted 2.6.9-55.ELsmp

What OS is your golden client running?  This looks like a EL4 kernel,
presumably taken from your client.

Cheers,

Bernard

--
SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada.
The future of the web can't happen without you.  Join us at MIX09 to help
pave the way to the Next Web now. Learn more and register at
http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] in.tftp not found in your PATH

Hi Adam:

On Mon, Dec 8, 2008 at 3:21 PM, Adam LaMee adamla...@gmail.com wrote:

 I'm getting an error when running si_mkbootserver:
 Looking for a tftp client... in.tftpd not found in your PATH.

What version of SystemImager?

 I'm using Ubuntu 8.10 and tftpd-hpa installed.  What is the expected path?
  in.tftpd is in /usr/sbin/ which seems to be typical.  Any ideas?  Thanks in
 advance,

http://trac.systemimager.org/browser/trunk/sbin/si_mkbootserver

644 echo -n Looking for a tftp client... 
645 TFTP_CLIENT=`which tftp`

It is looking for the tftp client, not the server.  The error message
is misleading and has already been fixed in trunk.

Cheers,

Bernard

--
SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada.
The future of the web can't happen without you.  Join us at MIX09 to help
pave the way to the Next Web now. Learn more and register at
http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] SystemImager based on a 2.4 Kernel

Hi Drew:

On Wed, Dec 10, 2008 at 10:39 AM, Drew Wrobel phoenix1...@hotmail.com wrote:

 What was the last stable version of systemimager that was based on a 2.4
 kernel?

Try 3.4.1.

Cheers,

Bernard

--
SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada.
The future of the web can't happen without you.  Join us at MIX09 to help
pave the way to the Next Web now. Learn more and register at
http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Help getting netboot working

Hi David:

Have a look at this thread, I believe you can do what you want:

http://www.mail-archive.com/sisuite-users@lists.sourceforge.net/msg04676.html

Basically what you need to do is add option-140 to your Windows DHCP
server with the ip address of your image server.

If you do get around to set this up successfully, it would be great if
you can add your experience to our Wiki:

http://wiki.systemimager.org

Thanks!

Bernard

On Fri, Dec 5, 2008 at 9:13 AM, David Young youngde...@gmail.com wrote:
 Hi. We're using System Imager 4.0.2. We have a rack of Sunfire servers upon
 which we're installing CentOS 5.2; a CentOS 5.2 virtual machine is our image
 server, and a Sunfire is the golden client.

 So. Using si_mkautoinstallcd works just fine, and we have two of our
 Sunfires installed this way. However, for the remaining machines I'd like to
 use a network install and am having trouble getting this to work.

 Our DHCP server is a Windows box (our sysadmin doesn't like this fact but
 that's the way things are right now), and again, our image server is a
 CentOS 5.2 VM. Questions:

 1. is it even possible to use netboot in our present environment (i.e. the
 image and dhcp servers are not co-resident)?
 2. if it is possible, then I'm obviously not understanding how to set this
 up; I ran si_mkbootserver --localdhcp=n, and it completes successfully; then
 I get these prompts:

 Once you're DHCP server is configured, you should be all set.
 Do you want to run si_mkdhcpserver to configure your DHCP server ([y]/n)? 

 I answered 'n' to the question since we already have a dhcp server running.
 My (admittedly limited) understand of dhcp is that only a single server may
 properly service a network; is that correct?

 Anyhow, booting one of the Sunfires from the network shows DHCP and a
 spinner; this runs for awhile then returns with no boot filename received,
 which I guess makes sense since our company dhcp server knows nothing about
 our image server.

 What do I need to do to get this working without replacing our dhcp server
 or moving our image server. Thanks.

 peace, david

 And now these three remain: faith, hope, and love.
 But the greatest of these is love.
  -- 1 Corinthians 13:13

 For wisdom is more precious than rubies,
 and nothing you desire can compare with her.
  -- Proverbs 8:11

 --
 SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada.
 The future of the web can't happen without you.  Join us at MIX09 to help
 pave the way to the Next Web now. Learn more and register at
 http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/
 ___
 sisuite-users mailing list
 sisuite-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/sisuite-users



--
SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada.
The future of the web can't happen without you.  Join us at MIX09 to help
pave the way to the Next Web now. Learn more and register at
http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] Problem with SI 4.0.2 and a SLES 8 client image

Hi Drew:

On Thu, Dec 4, 2008 at 11:59 AM, Drew Wrobel phoenix1...@hotmail.com wrote:

   But when the client reboots, the / partition does not report the correct
 size.

Can you please elaborate on this point?

   Would I be wrong in guessing that my problem is that a partition table
 created
 under a 2.6 kernel is formatted differently so that a 2.4 kernel doesn't
 read it properly?

I am thinking it is because the parted that came with SystemImager
4.0.2 created partitions which confused the 2.4 kernel.

I wonder if there are some compatibility flags that could be passed to
parted that can create partitions that does not cause problems for the
2.4. kernel -- also, have you tried updating the 2.4 kernel to the
latest version available?

Cheers,

Bernard

--
SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada.
The future of the web can't happen without you.  Join us at MIX09 to help
pave the way to the Next Web now. Learn more and register at
http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/
___
sisuite-users mailing list
sisuite-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-users

Re: [sisuite-users] rsync issue with image retrieval from golden client