Allow for retrying failed migrations
The service must allow for retrial (both automated and manual) of failed migrations.
Thank you for the feedback. Ability to restart/resume the failed migrations is in our review currently. We will share more details as we plan through this feature.
Jonathan Fite commented
I've developed a workaround that may prove helpful for others. It adds a step, but appears more reliable than uploading the file. I created a VM in Azure (simple, just a machine that can host a fileshare). I also installed SQL Express 2016 on there as well. So my process is:
* Create schemaless database on SQL Express (on VM in cloud) with the correct name (just to make checking the box easier)
* Create Share on VM in cloud
* robocopy backup file(s) (source and log) to share in cloud
* Create DMS project and activity, but use the VM in cloud (and SQL Express) as the "source" systems.
* Enjoy a fast and seemingly failure free network experience.
DMS appears to only want a SQL box to talk to in order to do HEADERONLY restores, so providing SQL Express is enough for this.
Worked for me from SQL 2016 to Azure Managed Instance.
JR Boyens commented
This was on the radar a year ago, is this still on the radar?
Masayuki Ozawa commented
I want to re-execute not only at the time of failure but also on the successful activity, in order to execute the same process again.
Thank you for the feedback. Ability to restart the failed migrations is definitely in our radar. I will let you know the approximately timelines soon.