Scheduled jobs drop from the scheduler needing a reload of the configuration file

I have many scheduled jobs that run smoothly without any issues but I am noticing that two of them randomly get dropped from the scheduler. The fix is that I just have to reload the configuration file from the File tab (so no changes to the files). Then rinse and repeat like a week later. I don't catch these until users tell me that they are not getting some of the expected reports. Has anyone encountered this and found a fix? Thanks

@jonatanencarnacion

Dear Jonathan,

Thank you very much for reaching out through the Devolutions Forum.

My name is Ruben Tapia, a support engineer reviewing request F-55574. To help us determine why these two scheduled jobs stop running until the configuration is reloaded, could you please provide the following information?

• PowerShell Universal version and build.
• Installation type, such as MSI, IIS, ZIP, or Docker.
• Whether this is a single-node or multi-node deployment.
• Database type being used.
• Whether the same two schedules are always affected.
• Whether the affected schedules remain visible under Automation > Schedules when the issue occurs.
• Whether they still appear in the .universal\schedules.ps1 file.
• Whether the Next Execution value is missing or no longer updating.
• Sanitized definitions of the affected schedules, including their schedule expressions and environments.
• Confirmation of whether they were created through the Admin Console, configuration files, or Git Sync.
• PSU logs covering the time when the schedules stopped running.
• Whether any Git synchronization, antivirus, backup software, or external process modifies files within the PSU repository.

Before reloading the configuration again, please capture screenshots of the affected schedules and their current Next Execution values. This will help us determine whether the schedule definitions remain present but are no longer registered by the scheduler.

Best regards,

Ruben Tapia

Hello Ruben,

-We are currently on 2026.2.2 but this has been happening with every version since we upgraded from PSU 4 to PSU 5 - Current Revision number: {12DE80BD-7AF5-46C7-8375-B178974C8BC4}
-MSI installation type
-Multi-Node
-MSSQL
-For the most part yes. Honestly do not know if other scripts get affected as they are not as high priority as these two.
-It does not. That is when we first noticed that it "drops" from the schedule
-Yes they are still on the schedules.ps1 file
-No next execution is visible as the script "drops" completely until a reload happens and then the next execution shows the next time as normal
-I am not clear on what you are asking in this part: Sanitized definitions of the affected schedules, including their schedule expressions and environments.
-the scheduled is created by updating the schedules.ps1 file. Not created directly through the admin console.
-We have git sync that syncs our repository between the various nodes. Nothing else modifies the files.

Hello Jonathan,

Thank you for providing these additional details. The fact that the affected schedules remain in `.universal\schedules.ps1` but disappear from the Admin Console and no longer show a Next Execution value helps narrow the issue to schedule registration rather than permanent removal from the repository.

By “sanitized schedule definitions,” we mean only the `New-PSUSchedule` blocks for the two affected schedules. You may replace internal script names, paths, parameters, server names, and other sensitive values with generic placeholders.

Could you also confirm:

* Whether every node is running PSU 2026.2.2.
* The total number of PSU nodes.
* Whether each node has an independent local repository clone or uses a shared repository path.
* Whether Git Sync is configured as one-way or two-way.
* Whether the schedule disappears from every node or only from a particular node when accessed directly without the load balancer.

When the issue occurs again, please avoid reloading the configuration immediately. First, check each node directly and capture:

* Whether the schedule is visible.
* The Current Revision displayed by PSU.
* The current Git commit.
* The hash and last modified time of `.universal\schedules.ps1`.
* Whether the schedule appears in Automation > Jobs > Diagnostics.
* Sanitized log excerpts from each node around the time the schedule disappeared.

Please do not post credentials, tokens, private hostnames, customer data, complete logs, or unredacted configuration files. Only include the relevant sanitized sections.

This comparison should help determine whether the issue affects all nodes consistently or is isolated to configuration processing on a specific node.

Best regards,
Ruben Tapia

The two scheduled definitions are as follow:
New-PSUSchedule -Cron "*/15 * * * *" -Script "PATH\Generic Script One Name.ps1" -TimeZone "US/Central" -Name "Generic Script One"
New-PSUSchedule -Cron "00 6 15 2,5,8,11 *" -Script "PATH\Generic Script Two Name.ps1" -TimeZone "US/Central" -Name "Generic Script Two"

I should mention here that we have other scheduled scripts following this same format and for the most part, we do not notice those being dropped.

Answer to questions:
-Every node is running 2026.2.2
-We have 3 nodes
-Every node has an independent local repository clone
-All nodes are configured one way
-This one I cannot fully confirm to be honest, but I can say that it looks like it disappears from every node

I will post a follow up when it happens again, it seems to be dropping weekly or every two weeks right now

Thanks

Hello jonatanencarnacion,

Thank you for the additional details.

Git Sync may be involved, but we do not yet have enough evidence to confirm that Git itself is removing the schedules. Since the definitions remain in `.universal\schedules.ps1` and reloading the file restores them, the issue may be occurring during schedule registration or configuration processing.

Could you please confirm:

- Which Git provider hosts the repository?
- Whether **Use External Git Client** is enabled.
- Where `schedules.ps1` is edited originally.
- The Git Sync interval.

When the issue happens again, before reloading the configuration, please compare the three nodes and capture:

- Current Revision and Git commit
- `git status --short`
- Hash of `.universal\schedules.ps1`
- Whether the schedules appear in Automation > Schedules
- Whether they appear in Automation > Jobs > Diagnostics
- Relevant PSU logs from the same time window

If all three nodes have the same commit and file hash while the schedules are missing, that would point more toward schedule registration than Git transport.

Because this is a public forum, please sanitize all logs, screenshots, paths, hostnames, IP addresses, usernames, tokens, and other environment details before posting.

Best regards,
Ruben Tapia

Hello jonatanencarnacion,

We are continuing the laboratory reproduction using PowerShell Universal 2026.2.2.

The initial single-node test confirmed that the schedules register correctly and retain their Next Execution values. We did not reproduce the disappearance in that scenario.

We are now preparing a closer match to your environment with three nodes, a shared database, and Git synchronization. This phase requires additional environment capacity and configuration before we can validate the multi-node behavior reliably.

At this time, there is no confirmed cause and no evidence that the schedule definitions are being removed from schedules.ps1.

We will share another update once the multi-node validation is complete.

Best regards,
Ruben Tapia