Failed to load resources - BadRequest error across all Azure AI Foundry projects

Hi Balaji, Krithika ,

That "Failed to load resources – BadRequest" you're seeing on every project's Overview page lines up almost exactly with another report from the same week, where the Overview page returned a BadRequest carrying NegativeMaxRowsNumber – "Max rows should be larger than 0." In that case it turned out to be a portal-side query regression, not anything wrong with the projects or their data — and it cleared on its own once the platform side was fixed.

In plain terms: the Overview page fires a backend query whose row limit is being calculated as zero or negative when the time window resolves to an empty interval, so the service rejects it with BadRequest. Your resources are fine underneath; it's the portal's read that's breaking.

Here's how to confirm that and keep working:

First, open the browser dev tools (F12) → Network tab and reload the Overview page. To verify: find the request that returns 400/BadRequest and check its Response body — if you see NegativeMaxRowsNumber or a metrics/rows message, you're on the same regression.

Next, read the same data outside Foundry — Azure portal → your Azure AI Services / Azure OpenAI resource → Metrics blade (or the Azure Monitor Metrics REST API). To verify: if the metrics/resources render here, the underlying data is intact and the fault is isolated to the Foundry portal.

Then, on any metrics view inside Foundry, switch the time range from the default to a custom window (e.g., a fixed past week). To verify: if data returns for some ranges but not the default, that confirms the time-range/row-limit edge case.

Finally, check Azure Service Health (not the public Status page) for a subscription/region-scoped Foundry advisory. To verify: Service Health will show any active incident tied to your subscription and give the authoritative status.

The time-range change and the Metrics-blade route are workarounds, not a permanent fix — the durable fix is server-side.

If it's still failing after you've confirmed data loads in the Metrics blade, paste the Response body of that red BadRequest request from the Network tab (with the four request IDs) and I'll help pin down whether it's the same platform regression or an RBAC/network scoping issue. Kindly let us know if the above helps or you need further assistance on this issue.

If the answer is helpful, please click "Accept Answer" and kindly upvote it. If you have extra questions about this answer, please click "Comment".