How can we increase job host shutdown timing in azure function ?

Question

How can we increase job host shutdown timing in azure function ?

Amar Agnihotri 926

Hi,
I am having a blob trigger which is getting fired but it is not completing its execution. When i checked the logs in Application Insights i found these

It is showing that my job host started running at 6:56 but started shutting down at 7:30. This might be the reason behind restricting the blob trigger getting completed. Can anyone help me to get a way of getting this done ?

Thanks

Anonymous

2022-10-27T14:44:28.707+00:00

The data-factory tag doesn't seem to apply here.

Regarding your problem: I see messages saying the blob trigger succeeded at 7:07:24.434. The entry right before this was an Exception which may be why the Function didn't accomplish its task but it actually completed.

1 answer

Your answer

Anonymous

2022-10-27T14:44:28.707+00:00

The data-factory tag doesn't seem to apply here.

Regarding your problem: I see messages saying the blob trigger succeeded at 7:07:24.434. The entry right before this was an Exception which may be why the Function didn't accomplish its task but it actually completed.

Answer 1

MayankBargali-MSFT 71,021 Moderator

@Amar Agnihotri Thanks for reaching out. As per the application insights screenshot you can see the error Cannot perform the runtime binding on a null reference so it looks like there is binding configuration issue on your function app code. As you are using storage binding, please refer to this document to validate the configuration for different supported languages and validate the same with your function code.

In case if you still need assistance, please share the sample code along with your binding configuration so I can assist you further.

Amar Agnihotri 926 Reputation points

2022-10-28T12:56:22.983+00:00

hi @MayankBargali-MSFT
Thanks for the reply. I have resolved that exception and my blob trigger is working fine when i am testing it from my local machine ( I have created this blob trigger in Visual Studio on my machine and then i published it to azure ). After publishing it is getting executed for some time but after it started showing host stopped and that's why i am not getting result because it is getting shut down in between. I have collected the traces and it s showing here

READING WHILE LOOP IS A MESSAGE DISPLAYED INSIDE THE CODE TO CHECK THE PROGRESS OF EXECUTION OF CODE. ( FOR THE TESTING PURPOSE)

My blob trigger script is attached as a text file . You can check that .<REDACTED>

Now this blob trigger is performing Upsert operation . As soon the new blob "Daily_Updated_Issues_With_ChangeLog_All_Projects.json" will be created as a result of adf pipeline run . It will compare the keys from this blob from the existing blob "All_Issues_With_ChangeLog_All_Projects.json" and will update the existing keys with updated entries and if key is not present in the existing blob then it will insert those keys . After performing Upsert operation it is writing all the data to new json file "All_Issues_Upsert.json".

If you will see the code you will easily understand the functionality. This is working fine when i am running it locally but the issue is coming when running from cloud .
Amar Agnihotri 926 Reputation points

2022-10-29T06:53:01.543+00:00

hi @MayankBargali-MSFT
I checked and found that i was using consumption plan and the timeout limit for Azure function in consumption plan is 10 min that's why the job host was shutting down when it was taking more than 10 minutes. I Now, I have changed my plan to premium one and function is taking little time for execution but now i am getting another error of CPU threshold limit exceeded and that's why now the host is getting stopped

Can you please tell me the settings to increase CPU threshold at run time ?
MayankBargali-MSFT 71,021 Reputation points Moderator

2022-10-31T09:20:28.21+00:00

@Amar Agnihotri Thanks for your response. If you look into the logs before the CPU threads exceeded, you can see there were out of memory exceptions, so it looks like you are doing CPU and memory intensive task. You can try configuring the 64 bit platform on your function app and see if it helps you. You can either move to the higher SKU or configure the auto scale on your function.

I will also suggest you to review your function code to troubleshoot the high CPU/memory consumption.
Amar Agnihotri 926 Reputation points

2022-10-31T16:32:18.31+00:00

hi @MayankBargali-MSFT
I made the above changes

But still it is getting stopped
MayankBargali-MSFT 71,021 Reputation points Moderator

2022-11-08T12:06:28.953+00:00

@Amar Agnihotri Apology for missing your last comment. In case if you still didn't find the root cause/need assistance I will suggest you to create a support ticket with us. If you don't have the support plan then please let me know so I can enable one time support request on your subscription.
Amar Agnihotri 926 Reputation points

2022-11-16T06:36:17.363+00:00

hi @MayankBargali-MSFT ,

The issue is resolved now. Actually, the json file was so gigantic that my azure function was not getting required computation power and that's why it was showing CPU threshold exception. When i scaled it up to EP2 then it is working fine. Thanks for your support and all the efforts.
MayankBargali-MSFT 71,021 Reputation points Moderator

2022-11-16T10:02:30.18+00:00

@Amar Agnihotri Thanks for your response and glad to know that the issue was resolved moving to higher SKU. Feel free to get back to us in Microsoft Q&A if you need any assistance in future.