Our remote workers use Direct Access to connect to applications and services hosted in Azure.

Outbound management from Azure to the clients is currently not support.
Microsoft have confirmed Azure does not support IPv6 over IPv4 - Protocol 41.

It would be a huge help if we could get support for this.

Only Account Administrator is currently allowed to Add Credit to Azure in Open subscription. It is extremely limiting and risky to have such low bus-factor.

It is impossible to change site recovery resynchronization cache size from 8GB.
If exceeding this 8GB, I want to change the resynchronization cache size to larger size.

When doing an update, the little bell icon has a running indicator to show that an update is in progress. But very often an update is connected to something going on in another browser tab - so I find myself quite often switching back to the Azure portal page to check if the update has finished.

If you add an indicator to the title element of the portal page that shows that there are still pending updates, that would prevent the constant flipping between tabs.

The limit of 256k for event hub and IoT messages is too small.

We have scenarios where we occasionally (0.1% of the messages) need to handle up to 1MB of json in a single message. Not being able to handle these renders event hub, and consequently Azure, unsuitable for purpose.

Kinesis supports up to 1MB messages, Kafka can support considerably more.
For Azure to be viable as a message endpoint the message size limit must be increased.

Customers excited about TSI, but they would like to deploy one in Japan instead of foreign countries

I was Kudu-ing a WebApp to observe some files and had a colleague of mine shutting down the same Webapp and i had no clue she was doing it. It would be nice if the logged in user accessing a specific resource is notified when the same is being modified by another Azure user

Message of logged exception is often cut off. Full data cannot be displayed.
Please increase length of message property to get full data.

With teams working all over the globe, payment should also be possible from any location. I had the same issue with Dell; however when we did a chat with them, they were able to accommodate payment at dell.ca with a USA credit card...So I'm surprised you don't offer payment from anywhere for Azure services. Companies are global, resources could be in X data center in X country, payment for said services should be payable from any countri.

It would be great if you can add couple of more fields in the activity log , old and new values. For example if someone is updating an alert defnition Azure is logging the event but not capturing the old and new values of the updated alert definition

The sales material on your site uses different naming conventions to the technical or management material available through the portal which is highly confusing. For example, a support technician had to explain to me that the 'Standard 1: Medium' instance of the virtual server I am on is actually the 'S2: Standard' instance I see when i go to the scale up page.

need a monitor metric about Memory percentage such as CPU percantage

Setting up an alert for a new VM or delete VM will send six emails as it's a long running operation and there is no way to suppress or limit to just one email. Microsoft case =117071616046005 thanks

At this moment, you cann't test Azure CDN Rule Engine before sending for approval, approval of new Rules takes up to 4 hours. Which make things very difficult.

Offer Azure Container Services for Microsoft Azure Germany region. Not offering it makes the Germany region semi-useless, since most modern applications come packaged as containers due to the obvious advantages of deployment and better scaling with micro service granularity.

How would one effectively implement a custom sign-in experience for multiple clients/customers while adhering to the DRY principle?

Requirements

-The look and feel needs to be customized to our design but the same* for all customers
-*The logo needs to be specific to the customer

https://stackoverflow.com/questions/45122685/branding-white-labeling-the-b2c-login/45155524#45155524

I suggest to have documentation created for adding DNS suffix for VMs on Model1. Ref #117062215932932

After adding DNS Suffix, the VM always goes into unresponsive status and has to be rebooted via portal. At time people just wait for the M to come up on its own but this never happens. It has to be manually rebooted.