If you are running any task (in this case update task) on 50 nodes then I will surely suggest you to enable randomization rather than invoking the task at all the nodes at a time.
If you are running the task on 50 nodes, there is chance that ePO server will miss some of the nodes.
Thanks for the reply, however, I do not think that randomization is the issue here. Even if I try to wake up a single agent using a force complete policy and task update, on one of the affected nodes, it still will not update.
Did you check Windows Firewall Settings?
We do not use Windows Firewall. It is disabled by GPo.
Strictly speaking the server doesn't push updates to the clients: the clients pull them from the server (or repositories.) I only mention this because it makes visualising what's going on easier
If you have a number of machines that are failing to update, you need to start your investigation on the affected machines. Check the agent log on the affected machines to make sure that they are trying to run the update task, and if they are but it is failing, are there any obvious errors logged?
If the task is running but failing, check the mcscript.log, again for any errors. You may find that the affected machines all have something in common.
Good point Joe. I will have a look at the logs and see what I can determine. Thanks.