Feature Reqeusts

[NickSmith]

Hello, I recently have been working on converting our patching process to using batchpatch. It is already much more reliable than what we had with SCCM by a long shot. I did some searching around the forums for these features or ways to do what I am hoping can be done, but didn’t find a whole lot of info.

I saw the one about “do not patch server Y until server X is done” yes, that would be a great feature for those of us in dc patching, or high availability systems patching, but I have that dealt with for now with ‘maintenance windows’ of lists I generated from our old sccm environment.

Two other features that would be neat to add if possible is a randomization option when scheduling a task. Scheduling the same task on an entire tab of 150 machines will kick off that task on all 150 at once. As you can imagine this puts quite a load on the host system. I have mitigated this issue with setting the max sessions limit to 50, and I seem to more or less get tabs of tasks finished within their windows. Just seems that being able to randomize start time up to “X minutes” when applying a scheduled task to a tab would be pretty handy.

The last one I have a need/use for is the ability to create more than one “Email notification”. We have a few users with special server needs that it would be really handy if I could create separate email notifications for. Currently I would have to use the same email notification and send task completion emails for all the servers to all the people. For instance Person A wants to know when app Y is done patching and Person B wants to know when app Z is done patching. Currently I would have to email both person A and person B about both app Y and app Z.

None of these are critical needs. The multiple email notification would probably be at the top of my list personally though.

Thanks!

Nick

[doug]

Hi Nick – It’s great to hear that BP is working so well vs what you had with SCCM! We love to hear that kind of stuff.

---

The “do not patch server Y until server X is done” functionality is already available in BatchPatch. Please see the following links for a description and tutorials. The second link is a tutorial for one possible way to use the feature but I think it will be helpful even if you are not dealing specifically with virtual machine guest/host sequences.

Advanced Multi-Row Queue Sequence

Virtual Machine Guest host Update and Reboot Sequence Automation

---

The task scheduler randomization option makes sense. We have on our to-do list already something very similar to what you are describing, but I’m not sure of when this will be available. That said, 150 machines is actually not that many for the system running BatchPatch, even when it’s just a workstation class machine. I mean… yes it’s a lot, but we have customers doing thousands all at once, if you can believe it. I have personally done several hundred at once running BP on a standard quad-core workstation with 8GB of RAM without any real issues. So, I think generally you should be fine scheduling all 150 tasks to execute at the same time. The worst thing that can happen is the GUI becomes a bit less responsive for a few minutes, particularly while the tasks are first being launched. However, the actual tasks should generally complete without issues.

Also, if you really still wanted to break it up and didn’t want to use the ‘concurrent thread max’ setting, you could schedule 50 machines for time X, 50 machines for time X+5, and 50 machines for time X=10 (or something along those lines, at least)

---

The email notifications can be customized/overridden on a per-row basis. Highlight the hosts that you want to override the default notification setting for, and select ‘Actions > Email notification > Override default email notification settings’

---

I hope this helps! Let me know if you have any other questions, comments, or problems.

-Doug

[NickSmith]

Ah, thank you for the response!

I did want to reply to the 150 machines at once bit. I understand it is not that many (over all we are closer to 1000ish total). However, due to other circumstances in our environment, there are a few commands I have to run before the patching actually begins. Sending a whole bunch of remote commands at once, even to as little as 50 rows at once does still bog the system down. However the actual process of patching the load is much lower. Basically at the start of a scheduled window, I bog the machine down with launching 5 remote commands across 50 systems at once, once that initial bogging is cleared up, it runs very smoothly. Once I get those issues in our environment worked out to where I don’t have to issue commands to make sure the server is using the correct SUS server and can just issue the patching command, I expect this issue to pretty much go away.

I had meant to reply and mention I found the override settings for the email, but I hadn’t found that bit about the first part of my post. Thank you much! I’ve since recommended it to other IT departments across campus who are having issues with their SCCM setups as well. Seems any IT group that does servers on campus is having this same issue and we all pretty much decided to stop pulling our hair out over it at about the same time 🙂

[doug]

Hi Nick – I’d be curious to get more detail from you about the “launching 5 remote commands across 50 systems at once.” What do you mean by this? How are you executing 5 commands at once? What are the commands? Which ‘Remote Command’ action in BatchPatch are you using? There normally should not be any “bogging” like you describe, so I’m hoping we can identify the exact problem so that we can then fix it. The more detail you are able to provide about the exact steps you are taking and the exact menu items you are clicking to launch this process, the more likely we’ll be able to reproduce the issue. If we cannot reproduce it, then it will be tough to fix.

Thanks!

-Doug

[Author]

Posts