Hey everyone, I’m having trouble with my Umbraco CMS project. It’s running version 13.5.1 and I’ve noticed it’s super slow when creating or editing nodes using the _contentService.
The weird thing is, it works fine when I use a local SQL Server in Docker. It takes about 300ms to create a node. But when I switch to an Azure SQL database or even a physical SQL Server, it takes between 3 to 6 seconds! That’s a huge difference.
Here’s a quick example of what I’m doing:
var timer = new Stopwatch();
timer.Start();
bool nodeCreated;
if (shouldPublish)
nodeCreated = _contentService.SaveAndPublish(newNode).Success;
else
nodeCreated = _contentService.Save(newNode).Success;
timer.Stop();
var timeTaken = timer.ElapsedMilliseconds;
I’ve tried different connection strings but nothing seems to help. Has anyone else run into this problem? Any tips on how to speed things up when using Umbraco with Azure SQL? I know there’ll be some slowdown because of network stuff, but this seems way too much. Any ideas?
yo, have u checked ur azure sql config? sometimes tweaking things like max degree of parallelism or cost threshold for parallelism can help. also, make sure ur not hitting any resource limits. might be worth looking at query plans to see if there’s any weird stuff goin on
I’ve encountered similar performance issues when migrating Umbraco to Azure SQL. One often overlooked factor is the database’s service tier and DTU/vCore allocation. Insufficient resources can significantly impact query execution times, especially during write operations.
Consider upgrading your Azure SQL tier or increasing the allocated DTUs. Also, ensure you have proper indexing on the tables frequently accessed during node creation.
Another strategy is to review your content type definitions. Complex property types or deep nesting may increase query complexity and slow down save operations. Simplifying your content structure might yield significant performance improvements.
hey there, whisperingwind. have u checked the network latency to azure? maybe there’s a bottleneck or routing issue. did u try azure’s query performance insights for any clues? what are your thoughts?