rockystout

@Adam Driscoll Thanks Adam! As for the download center, that is currently blocked for us, so we can't get to the links there. I'm trying to get an exception but that is uncertain right now.

@Adam Driscoll Hey Adam, can direct download links still be added to version announcements again? We have strict firewall rules in place and we can access downloads from direct cdn links but unable to access it from the download PSU link on the website.

No, I had to go with another solution, this makes datagrids unusable for a lot of what I need. I think this is a bug but I just haven’t touched this in awhile.. I don’t have any data or screenshots to even open an issue in github.

PSU actually supports running python scripts natively if you create a python environment to run the scripts in. This is helpful for scheduling and consolidating both Powershell and python scripts in one place. It doesn’t have as many features as ps though.

Product: PowerShell Universal
Version: 1.4.6

I thought I recall reading somewhere that there is a way to do it but I can’t seem to find it. Is there a way? We have python script executing and that works fine but It would be nice to be able to pass in arguments. So is there a way?

Product: PowerShell Universal
Version: 1.4.6

Curious if anyone else is experiencing this? Our last upgrade, computer groups were working before, and after the upgrade, they stopped. In the past we had issues where we found that Hangfire was never cleaned up and we had millions of phantom jobs being queued up and never running. Phantom being old schedules that had been deleted, old queues that no longer exist, etc. It feels like something isn’t being cleaned up in Hangfire tables. It required us having to wipe all of the Hangfire and PSU job related tables and computer groups would start working again. It’s happened again where computer groups stopped working and wiping tables doesn’t seem to have fixed it this time.

This type of things has happened in several disconnected environments and seems to pop up randomly. We do have a support ticket open but in trying to help troubleshoot i’m wondering if anyone else is experiencing anything like this? Maybe narrow down what might trigger it, etc.

Windows \ SQL Server

Thanks

We just did 4.5.x to 5.5.x. The big thing was moving from queues to computer groups and making sure that all schedules were set with the computer group instead. We have one entry point into our prod cluster (only one server can be modified with git pushing to the others). I shut them all down, made changes to the schedules, computer groups, etc, and restarted just the one. Once that was up I would upgrade each additional node.

The biggest headache was git.. because of the upgrade I had to manually copy the repository to every other node so the .git folder would be 100% in sync.

Glad you figured it out! I figured out the issue on my side I think, not why it was working and what triggered it to just stop, but the root had to do with hangfire.hash table not being cleaned up through version upgrades and old queues persist.

We are experiencing a sudden issue in all environments where if we use a computer group schedules won’t run those jobs. It queues in hangfire and doesn’t show up in PSU jobs. Setting schedules to run on the computer name fixes it. I’ll have to see if we have that specific error in logs as well.

I’ll bite

Can you expand on it being laggy? Just takes a long time compared to Git? Git is pretty quick.. I agree. But did it take significantly longer? Large environment?

thanks!

Don’t tempt me

Truthfully, I might be able to get something together in time though.. but I don’t know when. With it being newer I just can’t justify using it in prod… yet. I have seen people make enough mistakes with the git repo that required me to have to fix that this might be worth trying to prioritize.

Deployments are a new way to deploy code in PSU environments. PSU packages the system up in I believe nupkg file to deploy from one server to the other. It is supposed to have other benefits as well such as load ordering where github just restores files in no particular order.

docs.powershelluniversal.com

Looking at this option.. anyone using it? Good, bad, ugly? In multi-server environments (multiple servers in one environment, like prod)?

We currently have link all environments to Git repos but it can be a little flaky… the simplicity of deployments look nice. I currently don’t have a non-prod environment with more than one server to validate how a deploy would work in that scenario so anyone’s experiences would be nice.

And my understanding is that the loading of a new deployment in an environment has a much more structured order compared to Git?

Thanks!

We are on 5.5.4 and installed a 5.5.4 agent. I see now in the connections that yes there is a version column and the new agent shows the version. Every other client is not 5.5* yet so I had to filter to see the one agent that does display the version.

Thank you!

I may be missing something.. but is there a way to get the version of an agent programmatically?

I can’t even find it when looking at PSUAgent.exe details. We have a fairly large environment with different servers being in flux as we develop a new process that utilizes agents and we are having a hard time tracking and verifying installed agent versions. ie.. verifying that our process actually installed the upgraded version we told it to.

How are others handling this?

I know the card is already closed and I had updated it there, but turning Disable Auto Reload on, stopping the service, and cleaning up the duplicates in the scripts.ps1 file, did fix the issue for 5.5.4.

We did not, but I just set it and will monitor

I am seeing a very large number of duplicates in my scripts as well. I cleared them out yesterday, restarted, and scripts.ps1 didnt have duplicates but there were still several scripts duplicated in the GUI.

Just looked again today, and we have a very large number of duplicates in scripts.ps1 again. I think I worked with you on a similar issue sometime last year on an earlier version but the duplication stopped.

also.. Just upgraded from 5.5.3 to 5.5.4

Support email sent.. I included files form our environment so I didn’t want to open straight to Github

Not sure of your environment \ Network but network appliances can be set with maximum connection durations… if that is applicable.

That would be VERY useful!

OMG, I think I misread the documents… I had a reply all set to go and just saw that scriptpath is optional as I was checking myself. Well, then that solves my issue lol! I could swear that scriptpath was required

Thank you everyone, and with 5.4 that should fully lock down security!

They are the foundation of our new management system we are building for the several hundred servers we manage. They have been fantastic and allows a really distributed approach vs the monolith we have had for years.

The big challenge for us, and it looks like Adam just added a new feature to help to 5.4, is how we protect tokens and prevent users from being able to modify scripts. The token issue will be taken care of in 5.4, i’m still trying to solve how to prevent modification of scripts that might have to run under an elevated service account login.

We are even considering storing scripts inside of a secret store that we Invoke-Expression the value to load the module. That still leaves the agent.json config that could be modified to execute something else.

@Adam Driscoll I recall you saying that a ps1 file was needed because of (and forgive me if I remember this wrong) how modules and such need to load, like the agent couldn’t just receive a PS script and IEX it in memory, correct? I guess i’m checking my memory and wondering if it is possible, with maybe even limited use cases, that the PS script to execute could be sent with the payload from the server and executed from memory without the need of a PS1 file? That would solve just about all of our security concerns and really lock down distributed systems.

5.3.2 appears to have fixed another event hub issue I was having as well. It wasn’t auth related but it is still working at least.

I don’t have anything to help but I’ve used it with 5.3.0, 5.3.1, and 5.3.2 and token auth has been working for us using both administrator and another app specific token. If it helps we are using a SQL back end

We are struggling with this very thing. We will be deploying to a few hundred servers and we are trying to figure out how to protect the scripts as well as the service will have to run with elevated privileges. Im even considering keeping them on a NAS share so that even admins who get on the box wont have access. But the agents.json file must still exist so that still gives an avenue of modification. If you come up with a solution i’d love to hear it!