Baseline Reboots

(imported topic written by cstoneba)

Kind of confusing, but here goes nothing…

I have a task that reboots the client. The relevancy is if the BES client setting of reboot = true.

I then have this reboot task as a last component within our monthly patches baseline. Each patch baseline relevancy is put into the overall baseline relevancy by checking the checkbox for the baseline component. However, I am not checking the relevance checkbox for my reboot task.

My issue is when the patch fixlets are not applicable on a client, the baseline will not run (which is expected). The problem though is then the client doesn’t get rebooted, which we want to happen.

Basically I need a way to within my patching baseline to reboot the client if the reboot setting = true, else do nothing.

(imported comment written by BenKus)

In general, the problem is that you want the baseline to run but not show relevant, which isn’t really possible in any way that I can think of that you would like… How about make a separate Fixlet that restarts the computer?

You can make it with relevance like this:

client setting of reboot = true … AND uptime of operating system > 1 * day

Or something like that to prevent it from restarting the computer after the baseline restarts the computer… You can then target this Fixlet at the end of your Maintenance Window…


(imported comment written by cstoneba)

Yes, that would work, the downside is then we have another action running. It’s nice when it is all within a single baseline.

What processes do other users use for the automating patching & reboots?

(imported comment written by cstoneba)

anyone else?

(imported comment written by SystemAdmin)

An easy way to do that is to set a unique flag, time stamp, or registry value in your reboot task. Make both the task and baseline relevant when the flag, time stamp, or registry value doesn’t exist. That way the baseline will be applicable to and run on all the machines. The reboot will happen even if the other items aren’t relevant.

(imported comment written by cstoneba)

that’s the option i’ve thought about. The only downside is now the # of applicable computers count for the baseline will not be a true representation of the actual applicable fixlet count. But maybe that’s the trade off…