BigFix Client online duration

Content Authoring
1

Hi all,
I’m looking for some advice on checking the time that client has been online and able to download/install OS Patches or managed software.

In our environment we have very aggressive schedule when it comes down to OS Patching. We’re looking at anywhere between 93-97% of our machines that are online need to be patched in 8 days after Microsoft releases the patches (Patch Tuesday - Wednesday they are present in the console) clock starts ticking.

For machine to install patches following needs to happen

  • Patch needs to be relevant for the machine
  • Patch needs to be downloaded
  • patch needs to be installed
  • machine needs to reboot
  • machine needs to report updated status

Our environment is mix of Laptops/Desktops with majority being laptops, that are roaming between home/office/vpn/other custom locations. Our bigfix environment works both on the Internet and Intranet and currently running version 10.x

For the last few months we’re hitting easily 92-93% but getting to 95% and higher is proven to be challenging, and in checking some machines (Log files etc) we’re seeing that machines are not always online long enough to be able to download/install the patch. It’s time consuming to check logs on individual machines, so I’m looking for some advice (analysis, or other script) that could determine if machine is online, and able to download/install patch content. This is global customer and people can be on broadband (as fast as 1Gbps but also on 10Mbps and sometimes even lower) patches today are coming in at 500MB or more so it does take time to download and if connection is not stable retries etc. are needed.

I was thinking along the line of doing simple ping request to host that we now is online and available when machine is connected to corporate network, when machine is on the internet that’s more challenging but maybe could look at Adapter status, traffic going in/out. Ideally would like to analysis that populates property to let us know if machine has been online long enough so that Patch should be installed, this is very simplistic view, but I’m hoping you get the idea.

Any ideas/suggestions are welcome

2

The main problem I see is getting a machine that is offline, to tell you it’s offline. Often I’d use something in Web Reports to filter on “Last Report Time” to determine a machine is offline.

For “how long has it been online”, assuming that a powered-on machine should have some kind of Internet connection, you might create a property based on the ‘boot time of operating system’ or ‘uptime of operating system’ properties. Some examples

q: (now - boot time of operating system)
A: 01:13:20.250
T: 0.192 ms
I: singular time interval

q: (now - boot time of operating system) / minute
A: 73
T: 0.159 ms
I: singular integer

q: (now - boot time of operating system) / hour
A: 1
T: 0.128 ms
I: singular integer

q: uptime of operating system
A: 01:13:19.734
T: 0.096 ms
I: singular time interval


q: (now - boot time of operating system) / minute > 60
A: True
T: 0.068 ms
I: singular boolean

q: uptime of operating system / minute > 60
A: True
T: 0.033 ms
I: singular boolean

Your criteria then would be to find machines that are both “still online” based on a recent Last Report Time, and “have been online a while” based on your new custom property

3

Jason,
thx for the info. Today we already focus on machines that are online since we have triggered the patch action.
Example for July OS Patching.

Action was created 7/14. For our reporting we look at machines that have been online since 7/14.

The challenge still remains, while machine might have been online (last report time 7/14 or more recent) still does not guarantee machine has been online long enough to download/install/report back the patch status.

The check for uptime/boot time might be something we can add also, as for most companies nobody likes to reboot machine/device so it’s very common to have machines with uptime of 10+ days.

Will keep looking for solutions or options to determine online duration.

Thx for your insights and take care.