Unexpected AOS restarts on different environments

(3) Share

Report

Posted on by Diego Mancassola

862

Hi everyone, quick check with the community.

In the last couple of weeks I’ve been seeing unexpected AOS restarts on different environments.
Last week Microsoft helped resolve a stack overflow issue on one customer by enabling the flight EnabledStackOverflowPreventativeAction.

Now, on another customer, the flight is already active but AOS instances are still restarting. Microsoft mentioned it might be a platform-level session/runtime exception affecting multiple AOS instances simultaneously.

I also noticed there is an open quality issue in 10.0.46 (ID 1104344) related to intermittent crashes.

Before digging further with Microsoft, I’m curious:
Has anyone else experienced random AOS restarts or similar behavior recently?

Any feedback would be really helpful.

Categories:

Dynamics 365 Finance

I have the same question (0)

All responses (6)

Answers (0)

Sort by

André Arnaud de Cal... 304,688 Super User 2026 Season 1 on at

Like (0)

Report
Copy link

Link copied!

Hi Diego,

Recently, I haven't seen AOS restarts. The last time was July last year. That was related to huge memory consumption where if the available memory went down a specific threshold, the AOS instance got restarted.

Have you done some additional environment monitoring?

Was this reply helpful? Yes No
Diego Mancassola 862 on at

Like (1)

Report
Copy link

Link copied!

Hi André, tonight i will apply quality on Production.

I don't have any extra metrics to analyze the problem, the issue is currently investigation by Microsoft.

We try to find a correlation with SysMessage* framework that import data from production MES. We have the doubt there is something not working well here.

Was this reply helpful? Yes No
André Arnaud de Cal... 304,688 Super User 2026 Season 1 on at

Like (1)

Report
Copy link

Link copied!

Hi Diego,

Please update the thread once you know more. If it is related to memory, maybe an additional call of the finalize method on a class can help. Before doing that, check if this is useful and required or not.

Was this reply helpful? Yes No
Martina Kralj 14 on at

Like (0)

Report
Copy link

Link copied!

Hi Diego,

we’re seeing the same behaviour, but so far only on the LCS-provisioned Sandbox environments(v46).

Different AOS instances are crashing at different times. So some users can continue working in D365FO while others run into 502/503/504 errors on login or receive the “contact your system administrator” message.

We’ve escalated the issue to Microsoft and are waiting for them to enable the EnabledStackOverflowPreventativeAction flight to see whether it stabilises the environments.

At the moment, AOS services are crashing frequently due to stack-overflow errors. And we still can’t determine which specific change might have triggered this behaviour. We’ve also asked Microsoft to identify which call stack is causing the overflow so we can confirm whether any custom code is involved.
The issue started quite sporadically, which makes it even harder to pinpoint the root cause.

Best, M

Was this reply helpful? Yes No
Diego Mancassola 862 on at

Like (1)

Report
Copy link

Link copied!

Hello, thanks for your updating @Martina Kralj.

The flight for me resolve a specific problem where we loop code cased our bad data, and Microsoft can tell, from their telemetry, witch specific class cause the overflow. This is the official position of support: "platform‑level session/runtime exception affecting multiple AOS instances simultaneously. While the preventive flight helps mitigate certain scenarios, it does not fully eliminate all platform‑level crashes, which explains why the restart still occurred."

An update on my side, at the moment we don't have more restarts and the overall system is stable after installing last QU contains the issue 1104344 on 12/03

For now we don't have any official cause

We continue monitor for this week in case of new unexpected restart

Was this reply helpful? Yes No
Diego Mancassola 862 on at

Like (0)

Report
Copy link

Link copied!

Hello, after a week of installing the QU (cover issue 1104344) the system is stable.

We have closed support ticket.

Didn't have a certian correlation from Microsoft between the issue and our problem, but after installing it the system is stable.

Was this reply helpful? Yes No