This is related to dataset performance issues.
We are having serious issues with performance degradation today (09/09/2024) in East US.
We have premium P1 capacity, and no capacity constraints or throttling issues are appearing in the capacity metrics dashboard:
I checked the status page for Fabric, and they say everything is running smoothly:
https://support.fabric.microsoft.com/en-US/support/
However, it is clear that if we use SSMS to connect to any of our datasets in any of our workspaces that the service is not behaving well. We have been experiencing timeouts and disconnections all day long like so:
I will also be creating a support ticket, but just wanted to know if others had already done so. I would love to save some time, and I already have too many active support tickets .
Any information would be greatly appreciated. So far I've been testing the connectivity with these clients (.net):
Microsoft.AnalysisServices.AdomdClient 15.17.1
Microsoft.AnalysisServices.AdomdClient 19.84.1
and SSMS 19.2
Hi @lbendlin
I saw that too, but it doesn't correlate very well.
My problems are ongoing.
It might be a problem I need to review with our network team. They have started playing with lots of annoying components like zscaler security tunnels. Something like that may be interfering with the retrieval of data (for both my own machine, and for servers in the datacenter as well). At this point I think it is 50/50 that the problem is on our end or on Microsoft's.
The only reason it seems like a possible network problem is because in SQL profiler, the queries appear to finish in a normal amount of time like so (eg. ~2 seconds):
... but despite the fact that profiler says the query finished, we find the data is not being delivered. The TCP connection with Microsoft (20.42.6.81) isn't delivering packets to the SSMS application running locally on my workstation. It just hangs, and wait cursor in SSMS spins indefinitely.
Between my network team and my support case with Power BI, I hope to figure this out soon. Thank goodness for SQL profiler.
... Another weird thing I noticed is that if the queries return extremely small payloads then I don't get to the point where the network traffic will hang. But a moderately sized payload like 5 or 10 MB will cause everything to hang.
Hi @lbendlin
I saw that too, but it doesn't correlate very well.
My problems are ongoing.
It might be a problem I need to review with our network team. They have started playing with lots of annoying components like zscaler security tunnels. Something like that may be interfering with the retrieval of data (for both my own machine, and for servers in the datacenter as well). At this point I think it is 50/50 that the problem is on our end or on Microsoft's.
The only reason it seems like a possible network problem is because in SQL profiler, the queries appear to finish in a normal amount of time like so (eg. ~2 seconds):
... but despite the fact that profiler says the query finished, we find the data is not being delivered. The TCP connection with Microsoft (20.42.6.81) isn't delivering packets to the SSMS application running locally on my workstation. It just hangs, and wait cursor in SSMS spins indefinitely.
Between my network team and my support case with Power BI, I hope to figure this out soon. Thank goodness for SQL profiler.
... Another weird thing I noticed is that if the queries return extremely small payloads then I don't get to the point where the network traffic will hang. But a moderately sized payload like 5 or 10 MB will cause everything to hang.
Helpful resources
Microsoft Fabric & AI Learning Hackathon
Learn from experts, get hands-on experience, and win awesome prizes.
