Viewing 15 posts - 2,446 through 2,460 (of 22,224 total)
Time to employ extended events and capture the query metrics. That's the single best thing you can do. Identify how long each query takes, how frequently they are called,...
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
August 1, 2019 at 11:21 am
The same as with any performance problem, you have to gather metrics to identify where the issues lie. I can't tell you what the problem is because other than high...
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
August 1, 2019 at 11:16 am
That looks OK to me. Does it work?
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
July 31, 2019 at 12:40 pm
Check your databases. Sounds like one is corrupted. Run DBCC CHECKBD on them (during off hours, this is a fairly intensive query).
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
July 31, 2019 at 12:39 pm
First thing I'd do is understand if you're doing a lift & shift and just moving to Azure VMs, or, if you're implementing Azure SQL Databases or Azure Managed Instances....
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
July 30, 2019 at 11:03 am
That suggests a strong and distinct difference in I/O behavior. Are the different databases on the different instances on different disks?
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
July 30, 2019 at 10:51 am
Please, capture the wait statistics for the query on both fast and slow instances. Just running SELECT * over and over isn't going to tell us much unless we gather...
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
July 29, 2019 at 1:51 pm
So, the differences in index size and free space there suggests that the two databases are not identical. Was an index rebuild run on one of...
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
July 29, 2019 at 1:31 pm
small or large tables perform almost identical on Contoso
Ah, but "almost" is not identical. If everything is equal, performance should be identical (within a small margin of error, again,...
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
July 29, 2019 at 12:59 pm
Identical database. Check?
Identical query. Check?
If either of those is wrong, one database is older than another or something, statistics updates were run, whatever, then everything changes. However, assuming identical databases...
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
July 29, 2019 at 12:55 pm
But in SQL server 2014 I am not use compression
I bet it's configured there by default
run the query on both servers :
SELECT *
FROM sys.configurations
WHERE...
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
July 29, 2019 at 12:52 pm
The thing is, SELECT * from a table without a where clause is just a test of server settings, database settings, load, contention, and hardware. It's not anything else. There's...
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
July 29, 2019 at 12:49 pm
Did the recovery model get changed and the logs are now huge?
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
July 29, 2019 at 12:15 pm
Just piling on here.
When comparing the performance between two systems, we have to know that every possible thing is identical before we can say "See, SystemA is sucky and SystemB...
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
July 29, 2019 at 12:11 pm
Yes. If anything has changed in the structures such that the forced plan is no longer valid, it'll just get ignored.
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
July 29, 2019 at 12:07 pm
Viewing 15 posts - 2,446 through 2,460 (of 22,224 total)