Gateway Issues Impacting Power Automate and Power BI

(0) Share

Report

Posted on by razec18

119

Hello everyone,

I’m reaching out to the Power Automate community for assistance with a persistent Gateway issue that has been impacting our operations. When our Gateway crashes, all automations using SQL actions in Power Automate fail, which is critical for our workflows.

Issue Recap:

Initial Setup:
- Initially, we had a single VM with a gateway named PowerAutomateRPAProd. This VM and gateway managed all our Power Automate workflows and also updated all Power BI dashboards. Some of these dashboards are consumed by the RPA team, while others are used by business areas.
- Starting in August, we noticed intermittent Gateway crashes. RAM usage would spike to over 95%, and the Gateway service needed a manual restart to clear RAM and restore functionality. To mitigate this, we updated the Gateway to the latest version. This provided temporary relief, but the crashes soon resumed.
Scaling Resources:
- We increased the VM’s capacity to 32GB RAM and 8vCPU, exceeding the requirements for Gateway services. While this delayed the issue due to increased RAM capacity, eventually, all 32GB of RAM were consumed. Again, the Gateway service required manual restarts to clear RAM and restore normal operations, impacting both automations and dashboard refreshes.
New Gateway Deployment:
- We then deployed a new VM with a new Gateway (PowerBIRPAProd) to distribute the load. This Gateway was dedicated solely to BI dashboard updates.
- Initially, both Gateways functioned well. However, the issue has now reappeared with the PowerAutomateRPAProd Gateway, our primary workflow Gateway.

Recent Failures:

We encountered failures on December 23 and 24. Screenshots of these failures can be provided for reference if needed.

Observations & Questions:

Could anyone help identify why RAM consumption escalates this way?
Currently, the “Collect Additional Data” option is disabled. Should we enable this option to gather more information if the problem recurs?
Has anyone faced similar issues or have insights on potential causes and solutions?

Urgency:

This issue directly impacts our production automation environment, so finding a definitive resolution is crucial. Any guidance or recommendations from the community would be immensely appreciated.

Thank you for your help!

Categories:

Process Advisor

Screenshot_285.png

Screenshot_286.png

Screenshot_284.png

I have the same question (0)

All responses (4)

Answers (0)

lbendlin 8,363 Super User 2025 Season 2 on at

Like (0)

Report

Gateway Issues Impacting Power Automate and Power BI

Running a gateway on a single VM is risky. You are supposed to have two or more cluster members in each of your gateways. The Azure service will take care of the load balancing between the cluster members. DO NOT run anything else on these VMs.

32 GB of RAM is a good value. However, you can also restrict the memory usage in the gateway configuration - we set ours to 95% and haven't had any memory lockups since. Given multiple cluster members you can comfortable work with 16 GB of RAM on each cluster member.

For PowerPlatform requests you will want to have a good network pipe (10Gbps or better). For Power BI workloads in import mode you will want to have fast spindles, ideally separated from the OS drive. In both cases you want to have 8 or more cores per VM.

Run preemptive telemetry to make sure your gateways are healthy.

Was this reply helpful? Yes No
razec18 119 on at

Like (0)

Report

Gateway Issues Impacting Power Automate and Power BI

Thank you for your response! I understand that running the Gateway on a single VM is risky. In this case, to create a cluster, I would need the Recovery Key. Unfortunately, this Gateway was set up years ago by a team that is no longer with the company, and we don’t have the Recovery Key anymore. This means I would need to recreate all connections of Power Automate to set up a Gateway cluster with two machines.

Additionally, there’s nothing else running on these machines except the Gateway service, so that part is fine. What I find strange is that there are many Mashup Memory and Mashup services running continuously, and they don’t seem to stop. It’s as if they are accumulating RAM usage until the server crashes. When this happens, the Gateway either restarts itself after a while or I need to manually restart the Power Automate service for it to work again.

Regarding Power BI, as I mentioned, I’ve segregated it into a dedicated VM for the Gateway solely for Power BI, but I’m still facing issues with the Gateway and Power Automate machine. Do you have any ideas on how I could improve this situation? Or is there nothing that can be done?

Thanks again for your help!

Was this reply helpful? Yes No
lbendlin 8,363 Super User 2025 Season 2 on at

Like (0)

Report

Gateway Issues Impacting Power Automate and Power BI

If you don't know the recovery key you are doomed. Cut your losses and start from scratch.

Was this reply helpful? Yes No
razec18 119 on at

Like (0)

Report

Gateway Issues Impacting Power Automate and Power BI

There is no other solution?

To "recover" the recovery key, or maybe try another method?

Was this reply helpful? Yes No