This browser is no longer supported.
Upgrade to Microsoft Edge to take advantage of the latest features, security updates, and technical support.
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(144, 82%, 23%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EFC%3C/text%3E%3C/svg%3E)
Our team is building an application on top of SQL Server. We used to work with SQL Server Developer 2019, running on Docker. At that point, the queries run at an expected pace considering the hardware we have at our disposal: 12GB Ram, and a i5 11th generation Intel Core (4 cores, 2.4 GHz). That means, a 2 or 3 seconds max for big queries.
Last week we moved to SQL Server 2022 Developer edition and we moved Docker away. The server connects directly to the SQL Server on Windows. For small queries, the running became much faster, but for complex queries, much slower (in some case cases, 10 or more seconds for the same big queries). We notice that SQL Server is using less than 1GB ram (while in Docker we could assign around 4 GB). The processing capacity of the computer is also underused. The impression is that SQL Server is not using the full capacity of the hardware.
Why is that? According to the description of the SQL Server Developer edition, there should be no restrictions ...or does it have? How could we improve the performance?
Thanks
There are not very many hard facts in your post to give concrete answer. Let me first ask a few questions:
sp_configure 'max server memory' return?Generally, SQL Server grabs as much memory as it needs and as it can. So if SQL Server only allocates 1GB of memory, that would be one of:
Thanks a lot for your reply.
Agree, the queries are complex, but the workload does not need more memory.
Pity that you did not have Query Store enabled in SQL 2019. Then we could have dug out hard data about performance differences and investigated differences in query plans.
Since you are compat level 150, you have the same optimizer as in SQL 2019, so you are not a victim the changes in the optimizer back-firing. But there can still be plan differences for various reasons. (Of which not all are related to the version change.)
I guess then, we don't have much other options than to ask you to share a sample query that is slow together with the actual execution plan for that query. You can upload plans on http://www.pastetheplan.com.
Hi Erland, I found something interesting:
I uploaded the same database to Azure and picked a standard database for testing (10 DTUs)
"Complex" Queries take when running in the localhost up to 7 seconds for completion. In Azure, a fraction of a second. That is similar to what I experienced when running in Docker vs directly from Windows 11.
How could be that possible?
Did you compare the query plans? I would be surprised if they are the same.
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(128, 8%, 22%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EA%3C/text%3E%3C/svg%3E)
Hi @Fer C
There is a document on troubleshooting slow queries on SQL Server, which you may use as a reference: https://learn.microsoft.com/en-us/troubleshoot/sql/database-engine/performance/troubleshoot-slow-running-queries.
Best regards,
Aniya