What are the most common issues you run into with Automox?

The biggest ones for me;

1. The lack of realtime functionality. I get that normally it only scans on a set schedule based on the window you set but that is multiple hours in the low end. It kinda goes into a semi-realtime mode when you manually push a policy to an endpoint but I would like to see automox do something similar when you are drilled into a specific agent where all the data is refreshed in realtime. Typically when you are viewing an individual agent it is because something is wrong and you are needing to diagnose something. I know most other web based platforms can do this so its not a technical issue per se.
2. Related to #1 but I find somewhat regularly that the devices view is inaccurate compared to when you drill in (especially the status section). Would be nice is that is updated more frequently to make diagnosis easier.
3. I would personally like to see some thought be given to the worklets/policies and how they are displayed. I would like to see a single policy that can run across all 3 os’s (via some kind of a tabbed interface) instead of having to create 3 separate polices that each do the same function. Or maybe be smarter about OS detection and do not display mac policies/worklets on a pc or linux device and vice versa.
4. Tags; I would like to see the tag system expanded where you could create your own tag categories and use them separately. I personally want to log more data about my devices that is not pulled from the agent but only having one tag section makes it really difficult to parse out (since it exports all as one string).

Hi Aaron! Thanks for sharing these! Can you give me more specifics around what you’d want to display for the custom metadata from your devices (or what you’d put in TAGS) that is not already available on devices to sort/categorize? I’d love to hear more about your use case.

Asset numbers, physical location (room number and building info) are the things we currently have in the tags but there are others I would like to add that match our other systems. We have about 15 elements in our current inventory system and for now it would be good to mirror the data between multiple systems. The next step would be to tie it in with the api so you could have either automox populate info in our other systems or have those systems populate automox mapped tags. Then then allow us to decide which tags we want to see on the various views.

I could see that even being expanded beyond that data. A tag can be anything which will be different for each company.

We often see Automox agents becoming disconnected from Automox. Sometimes this is only for a few minutes before it reconnects, and other times it stays disconnected until we resolve the issue. We monitor the status via the API and receive alerts when this happens so we see that it happens across different organizations and environments (we’re an MSP) which leads me to believe the issue may be with Automox rather than the infrastructure. I’m curious to know if anyone else has noticed this in their Automox portals.

I’d also like to second Aaron’s first point.

In your cases, what does resolving it involve? Restarting the agent?

Yes, restarting the service resolves the issue 90% of the time I’d say.

Do you have a case open with support on this issue already? We’d love to get some logs to investigate further.

I don’t believe I’ve opened a support case for this exact issue. I will do so as soon as I get the chance though.

Thanks! Let us know whether they’re able to find out any root cause or fix for it.

• I have a large amount of machines in the pending reboot state. Would like to initiate their reboot from the console though give the user the same reboot notification experience they get from a patch policy job. Force the reboot to happen though ensure the user is apart of the process.

• Auto-Assign devices to groups based on Machine Name, IP, OS, domains, etc – Smart Groups

• Single post patch report indicating which machines failed to install updates with error code (exclude offline hosts). Allows for help-desk team to investigate.

Thanks for sharing! The auto-assign I’ve heard a few times from people.

@rleb great timing! I was just working on these requirements for Reboots notifications and it includes the “reboot from the console” item. We plan to begin coding this later in the year so please stay tuned!

Just want to stress here… very very very small sample size - I’m talking a couple days of light usage.

• I love love love the simplistic UI but I’d be nice to see some more detailed info on real-time patching and what’s currently goin on. I hit “patch” on something and basically just cross my fingers that something happens right now…

• More simplicity when it comes to the basic 101 stuff like pushing or uninstalling software. I know this is a patching platform but you have a huge opportunity to overtake products like PDQ if you’d just implement some simple yet effective software pushing capability that didn’t require some crazy worklet.

• I think I might be the only one but pushing an MSI which was injected with a key via Orca was a bit of an issue for me. My AD is sound and other software seems to push just fine but I’ve been running into a lot of issues getting the MSI with the access key injected to actually push successfully. This includes local and remote sites. I am going to dig into it more tomorrow and see if maybe there are some other tables that need modifying in Orca to streamline the silent install but today was mostly a manual process for me.

That all said… I am really REALLY digging how simple you guys have made this to be. I did my first couple patches and was like “uh… that’s it?” so that’s pretty badass. I think this has some serious potential to be a big big big solution for a lot of folks if you can iron out some of the smaller hiccups. Looking forward to learning it further and getting things all jiving.

Until the reboot functionality is introduced into the product, we do have a Windows Predictable Reboot Notification Worklet that will notify users a reboot is occurring at a time defined by the admin. You can send custom reboot notifications to the users at any frequency you desire.

Let me know if you need any assistance getting it setup.

Hi @dimforest, this great feedback so thank you for providing it! Some comments/questions for you in the order you listed them:

• Patch operation status: this is great feedback to receive and we’ve heard it before. It’s something we plan to tackle next year but are gathering information now. Can you define what real time means to you? How do you wish to see this information displayed?

• Install/Uninstall software, we will be working on this soon and will be releasing it next year. It would encompass an ability for you to install/uninstall software based on automation and approved/unapproved software. How and where would you like to see this information in the console?

• Any further information you can provide here from your digging would be helpful here.

I know you’re very active in the community and we love that, thank you! Please provide as much feedback, thoughts/ideas, ask us question and help us spread the word on this community. We will be able to offer a better product, faster for everyone from this communication.

@Greg thanks for the response! I have some more feedback and I plan to ping ya again in this topic when I get that all sorted upstairs in the noodle but for now, I’ll respond to your bullets:

• So right now, unless I’m mistaken or just overlooking it… I hit patch for a device and it brings me back up to the top of the page (which, btw… not a fan of. If I’m doing multiple manual patches it’s kind of a hinderance to scroll back down to try and figure out which one you just clicked, especially when there are a bunch and it doesn’t remove the one you just clicked to “patch now” from the listing right away) where I can see a spinning “loading” wheel. I’m not a UI guru but maybe if this was some sort of progress bar or percentage display, that’d give us more of a warm-and-fuzzy feeling knowing something is actually happening. If I have 3 things patching on a device, it’d be nice to click into that device and see 3 different progress bars or percentages showing where there are at in the process, if that makes sense. Furthermore, on the dashboard it’d be nice to parse all of this information from ALL devices so I could see all active patches being applied.

• This is a BIG one so I may devote an entire separate post to it but I’ll do high level right now. For starters, it should be it’s own tab on the left, maybe two. There’s still a ton of real estate over there so it won’t be crowded at all. Secondly, in my non-developer mind, I see it working a lot like ninite with the capability to add in custom packages, a la PDQ. The beauty of your platform though is the UI and the whole simplicity of it. PDQ is hosted 100% on-site and requires 100% on-site repositories for the deployable software… the second part I’m completely fine with but I like that Ninite has some of the more popular software “built-in” so I don’t need to maintain the files myself and can just click to install or deploy the latest and greatest. If you can maintain that level of simplicity but somehow still offer some of the more “advanced” functionality of PDQ… you’d turn this platform into a one-stop-shop software management solution for a ton of folks, myself included.

• MORE TO COME but for right now those are two of the bigger ones. Obviously I’m not privy to the vision and development so I may be suggesting things way off-base… but those were my initial impressions. I can’t stress enough how good looking the UI and layout is. It’s good to the point where you WANT to use it and maybe me wanting additional functionality is a part of me just looking for more excuses to be in it 😛

Honestly though, you have a great base solution here and even in the current state it’s worth the asking price. You also have a solid company behind it and that does go a long way nowadays. Really looking forward to see where you guys take this because like I said above, you have an opportunity to really carve out a solid spot in the market as being the one-stop-shop for software management and overall patching.

We definitely want to give the user more of that! I agree, adding more indications of various types of patching and system status is in the plan!

Good to hear! The main issue with the spinny wheel is it tells me nothing about progress and honestly… it reminds me of the “buffering” wheels elsewhere online and in other apps, which usually is associated with frustration, haha! At least if we had a progress bar or percentage (or both), we’d know that something is actually happening and could kind of gauge how long it’ll roughly take. Right now you click patch and watch a wheel for an unspecified amount of time wondering if you did it right.

Definitely heard your feedback always great to get that from users! We’ll keep you in mind or the Software management piece if we beta test it. Are you familiar with our public roadmap? If not, it’s here.

I’m familiar now!

Looks like it’s either not public or I screwed up registration and cannot access it…

Try this link:

That’s the ticket!

I find that the device status is often wrong or “stuck” in a mode. Sometimes for days/weeks its stuck at Scanning or Rebooting. Even though the endpoint is turned on, internet connected, and the automox service is running.

@fina27 has support worked with you on this? I want to make sure we get it documented so that we can find a root cause and get this fixed.

I haven’t created a ticket about it, no. Mainly because it doesn’t prevent it from working properly. Even if the status says Rebooting, I can still patch it successfully via the portal. And then usually it fixes itself. I’ll create a case once I get some extra time.

Thanks! Knowing the scope of the problem helps us prioritize fixes better, so that the most annoying bugs get squashed first.

I have seen that behavior on some endpoints too. Its not all devices but I would say at any given time 5% of the devices are showing incorrect status. Its hard to pin them down too since they do eventually go away (I assume there is a timeout on the backend but can’t really tell since we can’t see live status progress). Its even further to narrow down since you have to remember to go check the device multiple times to remember what mode it is in each time to see if it is truly not sending data back.

I think the command queue reporting feature would have greatly assisted with this but I see that idea has been pushed back further.

I view automox in its current form alot like SCCM, where you setup a policy/script and then you hope it applies but you really can’t tell in realtime what its doing. I know there are plans to address some of that and I am content to wait it out but it is an annoyance since it could be so much better.

Updated info on the agents on checking in issue. We do have a root cause for this one, and should have a fix out (hopefully) in the next version of the agent (not the one that came out today, the next one). It will be at least a month until that happens, given our current engineering schedule. In the meantime, support can help you get those statuses unstuck as needed if it’s happening to you and they don’t come unstuck on their own.

One point I missed responding to was around the ability to manually kick off a scan when you’re on a device page. There’s a Scan Device button in the upper right of that page, and that will force an immediate rescan rather than having to wait for the next group window. It’s also useful for testing the Evaluation portion of your worklets, as that code gets run when a device gets scanned.