mpagema

Product: PowerShell Universal
Version: 5.6.9

We push our scripts and schedules to production via Github. We are running 4 production servers. Whenever we push a new script+schedule to production. GitSync is configuration , PSU notifies about an error in the configuration

Invalid configuration: schedules.ps1

Script ‘.ps1’ not found

I belief this to be because the schedules.ps1 is reloaded, before the script file is in place. A reload of the configuration files fixes it. However, just on the 1 server I’m connected to. I’m having to do this on all the servers. That’s kinda silly for an automation tool, isn’t it?

Am I missing anything?

Product: PowerShell Universal
Version: 5.0.10

I’m developing my own module. As in /admin/platform/modules/MyModule

And while that’s working out fine, on pretty much a daily basis the code does not show. Thankfully the file on the server still contains all my code and it’s also still working. I can copy/paste it from the file and continue developing my module. But it’s rather annoying having to do that every day (sometime more often) I start working on it.

The modules directory is excluded from Git Sync.

Is this a know issue or something we can change in settings?

Looks good.

Having a PK won’t fix this issue; it’ll only speed up the logging.

(And I can see why making a PK would be an issue; the Id field contains duplicate values, so creating a PK will fail because of that)

Product: PowerShell Universal
Version: 4.2.15

The other day we had a long running job with loads of lines of logging to the pipeline. During the job, the transaction log grew to 750GB. While, the first 7 of 10 hours of the job, transaction log backups up to 250GB per hour were being made.

I’ve looked into the issue. PSU updates the joblog record for the running job on every new line sent to the pipeline. But since it can’t just add the new line, every new line causes an update to the existing joblog record, thereby also resending all logging already record previously.

I am aware of a work-around (collect the logging and send it in one go), but I’m wondering if a redesign of the logging method is a feature request that can be made.