This browser is no longer supported.
Upgrade to Microsoft Edge to take advantage of the latest features, security updates, and technical support.
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(89.60000000000001, 57.99999999999999%, 18%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EAN%3C/text%3E%3C/svg%3E)
Is there a way to use Azure Network Watcher or any other service to capture packets from an Azure Functions App? I have tried with Azure Network Watcher, but it only gives me an option to capture packets from a VM. The functions app is hosted on an App Service Plan.
@Asiful Nobel Can you provide more details on what other things you are looking for? You can use Azure Network Watcher capture creates capture sessions to track traffic in and out of virtual machines. As your function runs on the app service plan which has the dedicated VM therefore it will capture the session from VM's as mentioned in this document.
@Asiful Nobel Thanks for clarifying. In that case, you can use the Diagnose and solve problems option on your function app blade.
Diagnose and solve problems --> Diagnostic Tools --> Collect Network Trace --> Choose duration to collect the Network Trac and then click on "Collect on Network Trace"
Thanks for your help. I was able to collect trace from two Azure Functions app that I am hosting on two separate App Service plan. One is in an ASE v2 and one is not. Both app has the same code and configs. But the one running in ASE has delay connecting to the Azure SQL database and the other one has no delay. The delay is around 14 seconds.
This delay can be seen on below image. The 30.6 seconds here is the total API response duration and the ms durations are actual query execution durations. But the app is waiting a long time to start communicating with the DB. The DB is on Business Critical pricing tier with Elastic Pool.
The same API response can be seen with below durations on the app outside ASE.
So, I collected the Network Trace and it looks like the app outside ASE starts communicating with an IPv4 address starting with 40 immediately after getting HTTP request. But the app inside ASE does not start communicating with the same IP address until around 14 seconds pass by just like the App Insight transaction details show. During that time, the app keeps communicating with an IPv4 address starting with 52. Why might this be happening?
While investigating this, I found some interesting information about how Virtual Networks are setup for the ASE and the DB. It looks like ASE virtual network and DB virtual network is not the same. Can that be the cause? The non-ASE app has no virtual network configured.
@MayankBargali-MSFT Looks like I was measuring in an incorrect way. The ASE app had 5 instances. So, first request on each instance was taking around 14 seconds to open connection to Azure SQL Database. Subsequent requests are fast, probably due to pooled connection re-use.
Is there a way to reduce the duration of first connection open request?
@Asiful Nobel Ideally it should not have that much difference. If there is no request for 20 minutes then the app gets unloaded and the first request should again take time such as opening the connection etc. One of the reasons I could think is that you have enabled Always On for your function app where you see the first request is not taking time whereas on the other function it may be disabled.
Please also verify if there are no other requests coming to both of your function apps for at least 20 min to see if you observe the same behavior.
Please enable Always On in case if it is disabled and see if it helps you.