Just want to preface this by saying I’m fairly new to SysAdmin work. That said, I appreciate any insight and would kindly ask if replies could avoid assuming too much context. :)

I was tasked with opening a specific URL on every machine in the domain at a scheduled time. I set up a workflow (currently running only on manual starts from the Automation Hub) as follows:

1. Windows: Start-Process "explorer.exe" "[URL]"

2. macOS: open "[URL]"

3. Linux (Ubuntu 24 test machine): see below

On Windows and macOS, the commands work perfectly and the workflow ends as soon as the browser launches.

On Linux, however, I ran into issues. The Pulseway agent always seems to execute commands as root (even when the “Run as logged in user” option is selected).

I tried running the command as the seat0 user with runuser -u, but Firefox/xdg-open wasn’t inheriting $DISPLAY/$WAYLAND_DISPLAY, $XDG_RUNTIME_DIR, and $DBUS_SESSION_BUS_ADDRESS correctly.

After a lot of band-aid fixing, I ended up with this super scuffed one-liner (I am aware this is probably very suboptimal, but it at least launches Firefox so I'm not touching it until I know more):

U=$(loginctl list-sessions --no-legend | awk '$4=="seat0" && $3!="gdm" && $3!="lightdm"{print $3; exit}'); if [ -n "$U" ]; then UID=$(id -u "$U");
(runuser -u "$U" -- bash -c '((
env WAYLAND_DISPLAY=wayland-0 
XDG_RUNTIME_DIR=/run/user/'"$UID"' 
DBUS_SESSION_BUS_ADDRESS=unix:path=/run/user/'"$UID"'/bus
/usr/bin/firefox --no-remote --new-instance --new-window --start-maximized "[URL]"
>/dev/null 2>&1) &) & disown'); fi; exit 0

This does successfully launch Firefox with the URL, but the workflow action hangs for ~30 minutes (until max execution time hits, currently set to 1h).

I’ve tried:

• >/dev/null 2>&1 &

• nohup

• disown

• spawning a subshell and exiting it in hopes that would trigger the pulseway agent to get a kind of "command executed" signal.

But nothing has worked, the workflow still hangs instead of completing once the command is handed off.

So I think I’m misunderstanding something fundamental either about Linux process handling, or how Pulseway agents execute/track Linux commands within workflows.
Happy to provide any further information if needed, thank you in advance!

Veljko (Max)

Do you guys know of a way to make Pulseway send out notifications for an offline device that has been down every 3 hours or so? The initial notification is handy, but can we forget to go back to it. A 3-hour notification seems to be a good reminder to go fix the stuff. The other solution is sending it out to a ticketing system but that's just additional step that we would like to avoid.

I am wondering why the Context (for example Scope or Device) is only available when using the Ad hoc trigger type in a workflow, but not when selecting Notification as the trigger.

I am aware that I can add checks for the device inside the workflow itself, but I would prefer to execute the workflow only when a notification is triggered for specific device or devices. Is there a reason why this context information is not available at the trigger level?

I am not bothered when even a single device is online, I only want to know when all of them are offline.

cheers

pav

I suspect that this is likely due to the fact that I work remotely from Wisconsin (in the CST Time Zone), while my employer and many of our customers are located in California (in the PST Timezone).

For instance, I set the Schedule to Run every 2nd Thursday at 12:00 AM CST (the equivalent of Wednesday 10:00 PM PST) and a week or two later, I find that it is running a full 24 hours behind schedule.

That said, you may want to look into this, as the Schedules definitely don't seem to be functioning reliably, across Timezones, etc.

I am trying to run a workflow to automatically install an application via Workflows -> Create from Template -> Deploy Application From EXE File.
I have replaced the "Get URL" Workflow with "Write File" to download from the Managed Files storage area as described in the template.
However, when running the workflow it fails at the process of installation.
When I check the workflow history I see the following error: Timeout to respond with EXECUTE_FILE for 62bfb72f-9bff-433c-ba9a-426f17d5721a
Whenever I try this with another file (Notepad++), also stored in the Managed Files storage area, it works like a charm.

Any ideas?

Thanks in advance!

I've cobbled together a workflow using Powershell that looks at membership of the local Administrators group and that output is emailed to me. That's gone without issue but I want to refine it and that's where I'm struggling.

So I know that in our case, every machine's local admin group has some membership, i.e. the local admin account and our domain admins group. I want to filter these machines out and have email results for only those where there's anything more than those "defaults".

Get-LocalGroupMember -Group 'Administrators' | Where-Object {$_.PrincipalSource -ne 'Local' -and $_.Name -ne 'MyDomain\Administrator' -and $_.Name -ne 'MyDomain\Domain Admins'} | Select-Object -Property 'Name','PrincipalSource'

This code works locally and using Pulseway's remote Powershell function. What I was hoping to do was to basically skip a system if the output of the code was blank or null. For systems where I haven't received any output back when using Pulseway's Powershell, I guess I don't know if the output is actually completely empty or not. I also don't know if using Pulseway's variable function is considered separately to the Powershell environment the code gets run in, if that makes sense? When I try to run the workflow on a single machine I always get emailed the results, even when the return from the Powershell is blank.

I'm not brilliant with Powershell so have been searching around for alternatives. I was wondering about return codes perhaps but could I also put these into a Pulseway workflow?

Been looking around, but can't seem to find any examples - which a bit of a surprise to be honest.... 

Thanks!

Please advise. 

Thoughts?

I'm Ryan and a happy user of Pulseway, but kind of a starter with probably a little issue which I cannot solve. Could someone help me with the following case?

I provide customers with proactive automation on their devices and there is a issue which after a lot troubleshooting I still cannot fix. I run a policy which monitors daily statuses that check every single process and services. I got this issue ''The antivirus Windows Defender has been disabled on computer ...... in group ...... ''. The service stops daily and I made a workflow to start is again and also tried a workflow which restart the service, but in both cases the windows defender got the same error the next day (when a windows patch took place). There is no 3rd party anti virus in this case.

Could someone have an idea which I can try?

Kindly looking forward to a reply.

Regards,

Ryan. K

I've been looking all over and I can't find what are these "Action Parameters" in the "Condition":

Can someone explain how do these options work, please?