When RAIDs are used, metafiles are stored on the hard disks. This can result in a new installation not being able to run correctly and the operating system installer waits for manual input.
Using the "Pre-run NOC-PS Profile" function, another NOC-PS profile can be executed before the actual operating system installation.

The following should be noted:

  • The pre-run function is only executed on dedicated servers. VPS are excluded because VPS usually have only one hard disk.
  • If you enable this option, make sure that the "System Recovery Operating Systems" option is configured correctly. The pre-run profile will not run when the customer starts a recovery system.
    Incorrect configuration may result in data loss.
  • It is recommended not to use the default "wipe hard disk" profile of NOC-PS, because the execution takes a long time. Below is an example of a quick wipe custom profile.
  • Enabling this feature will increase the installation time, as 2 profiles must now be executed.
  • The profile will also run on newly assigned servers if you have enabled automatic operating system installation on delivery. 

How to enable the option

Click on "Addons" → "NOC-PS For WHMCS" → "Settings". On the settings page you will find the setting "Pre-run NOC-PS Profile".
Select the desired NOC-PS profile in the dropdown. Further steps are not necessary.






Example NOC-PS Profile

As mentioned above, the use of the standard NOC-PS profile is not recommended.

We have prepared the following example, which only overwrites the first few MB on each disk and additionally executes wipefs.
Fields that are not listed are left blank.

General settings:

PXE script:

#!ipxe
kernel http://$server/proxy.php/isolinux/rescue64 dodhcp rootpass=$rootpassword vncserver=1:$rootpassword boothttp=http://$server/proxy.php/sysrcd.dat setkmap=us ar_source=http://$server/kickstart.php/sysrcd?f= autoruns=no ar_nowait
initrd http://$server/proxy.php/isolinux/initram.igz
boot

Post-installation:

for i in $(lshw -class disk -class tape | grep 'logical name' | awk '{ print $NF }' | grep 'sd\|nvm'); do
    timeout 10 dd if=/dev/zero of=$i bs=20M count=1 &
done

sleep 15

for i in $(lshw -class disk -class tape | grep 'logical name' | awk '{ print $NF }' | grep 'sd\|nvm'); do
    wipefs -a -f $i
done

halt

You can customize the commands as you like.


  • No labels