Forum Widgets
Latest Discussions
Weird problem when comparing the answers from chat playground and answer from api
I'm running into a weird issue with Azure AI Foundry (gpt-4o-mini) and need help. I'm building a chatbot that classifies each user message into: follow-up to previous message repeat of an earlier message brand-new query The classification logic works perfectly in the Azure AI Foundry Chat Playground. But when I use the exact same prompt in Python via: AzureChatOpenAI() (LangChain) or the official Azure OpenAI code from "View Code" (client.chat.completions.create()) …I get totally different and often wrong results. I’ve already verified: same deployment name (gpt-4o-mini) same temperature / top_p / max_tokens same system and user messages even tried copy-pasting the full system prompt from the Playground But the API version still behaves very differently. It feels like Azure AI Foundry’s Chat Playground is using some kind of hidden system prompt, invisible scaffolding, or extra formatting that is NOT shown in the UI and NOT included in the “View Code” snippet. The Playground output is consistently more accurate than the raw API call. Question: Does the Chat Playground apply hidden instructions or pre-processing that we can’t see? And is there any way to: view those hidden prompts, or replicate Playground behavior exactly through the API or LangChain? If anyone has run into this or knows how to get identical behavior outside the Playground, I’d really appreciate the help.
Container Apps Environment Networking (Consumption)
AI workloads are no longer just about models, they’re about how those models connect. On Azure, many teams are running inference APIs, background processors, and event-driven AI components on Azure Container Apps Environments (CAE). CAE fits AI workloads well: it scales fast, scales to zero, and removes Kubernetes overhead. But once AI services need to securely reach private data sources, on-prem systems, vector databases, or external AI services, networking becomes the real design challenge. I’ve written a short, practical deep dive on how Consumption-based vs Workload-profile Container Environments behave from a networking perspective, what works, what doesn’t, and why it matters for modern AI platforms. 👉 Read the full article here: https://vakhsha.com/blog.html?post=blog-06Issue with gMSA when installing Cloud Sync
We are trying to install Cloud Sync to make use of the group writeback. However, we get the same error message every time we try to complete the installation We already tried: created a new sync server from scratch test the service account with "test-ADServiceAccount" check the encryption settings of the GMSA (the account is being created in the AD) removed an old orphaned GC tried it with a custom GMSA (same error) gave the server access to the GMSA via set-ADServiceAccount Did anyone else ever had this problem or know how to fix it?
Unable to delete Foundry Agent identity Entra app in Azure
I'm trying to delete an Entra app in Azure created by Foundry Agent identity blueprint as its currently unused and is causing EntraID hygiene alerts. However getting an error mentioning that delete is not supported. Is there any other way to delete an unused Entra app for an agent identity blueprint? Error detail: Agent Blueprints are not supported on the API version used in this request.
Understanding Storage Account replication downtime
I have a Storage account that's used as a CDN to host a lot of generally small files which occupy about 2GB. This is a small but critical part of our application which is used heavily by our app but which has no redundancy (it currently only has LRS replication). It's hosted in UK South and while Storage Accounts are very reliable, I'm concerned that if there's ever a regional outage there's nothing I'd be able to do. The requirements therefore are: Convert it from LRS to GZRS i.e. actively replicating from UK South to UK West. No app changes required to detect when the primary goes down and to switch to the secondary-this needs to be transparent. No or low downtime when the change is made. We need to be able to write to the secondary after failover. As a software company anything that limits our ability to push code changes is not acceptable, so RA-GZRS is off the table. After doing a bit of reading, I found the following warning in the docs: If you choose to perform a manual migration, downtime is required but you have more control over the timing of the migration process. https://learn.microsoft.com/en-us/azure/storage/common/redundancy-migration?tabs=portal#downtime-requirements This is typically light on detail and leaves some critical questions unanswered: Is there any way of estimating how long the downtime will be so I can appropriately set expectations of management and customers when scheduling the maintenance window needed? It specifically mentions manual migrations i.e. making the change through the Azure Portal, would making the change through IAC e.g. Bicep or Terraform be any different? Any input from anyone who's made any similar changes will also be appreciated. Edit: I've just checked and found that UK West still doesn't have Availability Zone support, is my best option for reducing the risk of this single point of failure to set the replication to GRS? https://learn.microsoft.com/en-us/azure/reliability/regions-list#azure-regions-list-1
How to troubleshoot if a cookie is being sent to application gateway with each and every request
I have a rule on WAF policy associated with application gateway with a rule (set as topmost rule) to allow traffic if a particular cookie is sent with the request. But we are seeing some requests that are not hitting that rule and instead hitting different rule and thus getting blocked. My thinking is that the cookie is not being sent by the application in that request, although the developer says that it should be sent with each request. How can I log enough detail on application gateway to see if a cookie was really sent with the request that was blocked or not.
