Killing a stuck sr-scan when vdi gets suck in vdi_deactivate

Simple fix for this one, kill the task directly by it’s PID. Using ps , grep and sed, to select correct process.

When executing this in live production, take especial care.

# xe task-list 
uuid ( RO)                : 7cb8b30b-563c-db9e-7682-e25aac2b2f07
          name-label ( RO): SR.scan
    name-description ( RO): 
              status ( RO): pending
            progress ( RO): 0.000
[root@21-16-246-494694 ~]# date
Thu Nov 10 11:41:45 UTC 2016
[root@xenhost ~]# kill -TERM $(ps fauxww | grep "[t]ap-ctl close" | sed -e "s/.*-p //" | awk '{print $1}')

Disabling SELinux

Today we had a customer that needed to perform a first generation server to next generation migration however they cannot have SELinux enabled during this process.

I explain to the customer how to disable this, it’s pretty simple.

vi /etc/sysconfig/selinux


Needs to be changed to


Job done. A simple one but nonetheless important stuff. If you wanted to automate this it would look something like this;

sed -i 's/SELINUX=enforcing/SELINUX=disabled/g'/etc/selinux/config

This sed oneliner simply swaps SELINUX=enforcing to SELINUX=disabled, pretty simple stuffs. It will work on CentOS6 and 7 anyway, and should but I can’t guarantee work on CentOS 5.