Automate Monthly AD Sync and Update Install w/reboot

[bleonard]

I’m looking for advice on how to automate our BatchPatch Windows update process. Ideally I would like to setup a recurring job that syncs with AD to add or remove new or deleted machine accounts. Then run a Windows update and reboot task only installing updates older than 30 days. At the end of the job sending out an email of the job status. I can do all these tasks individually, but I’m having trouble putting them into a single job that automatically runs every month. Any help is greatly appreciated.

[doug]

1. Configure grid sync settings (Grid > Synchronize grid with directory)

2. Configure 30-day threshold for Windows Updates (Tools > Settings > Windows Update > Only install updates that were published / approved at least 30 days ago)

3. Configure email notifications settings (Tools > Settings > Email notifications)

4. Create and save a job queue with steps that looks something like this:

—-A. Synchronize grid with directory (add and remove hosts)

—-B. Download and install updates plus reboot if required

—-C. Send email notification

Job Queue tutorial: https://batchpatch.com/using-the-job-queue-in-batchpatch-for-multi-step-execution

5. Create a recurring scheduled task that executes the job queue that you created/saved:

using-the-task-scheduler-in-batchpatch

creating-a-recurring-scheduled-task-in-batchpatch

[bleonard]

I believe I got it set up correctly. Is there a way to limit how many hosts BatchPatch is processing at once? Some of our OUs have 50+ servers in them. They are VMs and having too many VMs rebooting at once can cause an issue with the VM host machine.

[doug]

‘Tools > Settings > General > Concurrent Thread Maximum’ can be used to control the number of simultaneous threads. However, in your situation since you are scheduling the jobs I would recommend that you simply schedule the virtual hosts in a staggered fashion so that they execute at different times rather than trying to execute them all at the same time.

Another option is to use the multi row queue sequence to stagger execution:

advanced-multi-row-queue-sequence-video-tutorial

advanced-multi-row-queue-sequence

virtual-machine-guest-host-update-and-reboot-sequence-automation

[bleonard]

I was actually going to run the job by creating a new tab adding a row named after the job queue, then adding the recurring scheduled task pointing to the desired saved job queue. I noticed the option of “Wait for host to go offline and come back online” so I added that just before the Send email notification step. Am I correct in thinking this will wait until all 80 or so hosts have completed their reboot before it sends out the grid email? I want to avoid seeing hosts that are mid reboot or not complete with the install on the email. I’ve had issues in the past not giving BatchPatch enough time to complete before sending the email.

[doug]

“Wait for host to go offline and come back online” is explained in more detail here:

understanding-the-special-items-in-the-batchpatch-job-queue

[Author]

Posts