APIM backup cause restart

Question

Hi!

At my project we are migrating from one CSP to another. This means that we also have to migrate our APIM portals.
To do this I will use the Azure Powershell command "Backup-AzApiManagement" to create a backup to storage and restore this again on the new instance (with a different name in the new subscription) with the command "Restore-AzApiManagement".

This morning I started the backup, when I tried to access the developer portal I got a 404 not found. Apparently with the backup I triggered a restart:

---

Unavailable : Configuring (Customer Initiated)

At Friday, November 6, 2020, 11:21:01 AM GMT+1, the Azure monitoring system received the following information regarding your API Management service:
A configuration change is currently being applied to your API Management instance. A user-requested configuration change (e. g. setting up custom domains, virtual networks, moving resources or applying custom settings) is being applied and it requires restarting your instances.

Recommended Steps

• During a period of up to 45 minutes your service may be unavailable. If the service is not available after this period, please contact support .
• If you are experiencing problems you believe are caused by Azure, contact support

---

The APIM instance I am backing up is in the pricing tier "Developer (No SLA)", so I can imagine that this is the reason that the instance is restarting.

My questions:

• Is it normal for APIM to restart when creating a backup?
• Is the restart depending on the pricing tier?
• Is this behavior somewhere documented?

Thank you!
Kind regards

Answer 1

Hi @JorneS

APIM backup should not cause restart of the APIM service. As you have mentioned that you were getting a 404 error and 404 generally occurs when the requested URL is not correct. If there could be an issue with APIM so the error code should be 5XX.

As per the above log, I can see that the restart event was Customer Initiated and you have also confirmed that you have restarted the APIM service.
Unavailable: Configuring (Customer Initiated)

I have tested the scenario at my end for Developer SKU and couldn't observe a 404 error while accessing the developer portal when I have initiated the backup APIM call using Powershell. Further, I don't see a restart or any issue when the backup was initiated for other SKUs.

Please note the Developer tier is for non-production use cases and evaluations. It does not offer SLA.
You can refer to the APIM feature comparison for different SKU: https://learn.microsoft.com/en-us/azure/api-management/api-management-features

You can always reach out to us if you are facing any issue with your APIM resource. Please let me know if you have any queries or concerns.

Please 'Accept as answer' and ‘Upvote’ if it helped so that it can help others in the community looking for help on similar topics.

Answer 2

Hi @MayankBargali-MSFT

Good to read that a backup should not cause a restart of APIM services. I am aware that the 404 means HTTP NOT FOUND, my route was correct since I went to the developer portal via the button in the Azure Portal (also I went to the root).

I have not confirmed and have not restarted the instance, I asked the question here to get information and clarity. The only action I took was to initiate a backup with Azure Powershell, then the restart happened. At the time period the only action that was executed was the backup.

The APIM instance is used for testing purposes, so no production use cases. I am aware of the differences of the pricing tiers and I know that the developer portal can go offline for maintenance etc. because of the SLA.

My question here is to know if it could be normal behavior that a backup restarts APIM. Now I know this isn't the case.

Sorry, I am not going to upvote your response, you missed key parts of my question and answered it wrongly:

• I didn't restart the service
• I didn't confirm the restart
• You tested the 404 while the portal was working, my portal was restarting and then I got a 404.

Kind regards.

Answer 3

As you may already know, this is what happens when running backup on the APIM instance (If the SKU is developer). The API's works as expected but you cannot reach or make changes to the developer portal. Also the activivity log will say that the Health Status has changed.
I filed a case at Microsoft about this but because of the SLA of the SKU , they just said like "works as expected".

But the good news is that the standard or premiun tier does not behave like that.