December 1, 2025 at 3:50 pm
Hello,
I am receiving multiple direction on deploying SQL Servers in AWS EC2. I've always separated SQL Server data and log files. I then separate the TempDB files onto different drives. A few times, I used the log file drive for database log files and TempDB log files.
Now, in an AWS Ec2 environment, do all files go on the same drive? What is best practice on where we place SQL Server files?
Just trying to learn the AWS environment and best practice to manage SQL Server environment in AWS.
Thanks.
Things will work out. Get back up, change some parameters and recode.
December 2, 2025 at 4:10 pm
Thanks for posting your issue and hopefully someone will answer soon.
This is an automated bump to increase visibility of your question.
December 3, 2025 at 5:38 pm
Yeah, I'd still separate them. Same underlying issues are still there. Different levels and types of IO for each mechanism. Splitting them even lets you pick different levels of IOPs for each type of storage, saving money on, let's say logs, while you spend more money on tempdb. It's a great way to go.
"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
December 3, 2025 at 9:39 pm
I'd agree. In the cloud you often have software defined storage, so having more paths and being able to upgrade one is helpful.
December 4, 2025 at 5:07 pm
Thank you guys for that. I was thinking that, but this confirmation helps.
Things will work out. Get back up, change some parameters and recode.
February 3, 2026 at 12:05 pm
Yes even on AWS EC2, it’s best to separate data files, log files and TempDB into different volumes. They have different I/O patterns, and isolating TempDB in particular prevents it from becoming a bottleneck. This setup allows you to tune disk performance for each type and avoids contention.
March 25, 2026 at 10:06 pm
Personally, I'd separate the user database data and log files into separate volumes.
As for the tempdb database, AWS EC2 uses solid state drives (SSD) so there's no need to create separate volumes. Both your TEMPDB data and log files as well as all other system databases can reside on the same volume. These drives aren't your old school spinning disks lol.
What's more important is having the number of tempdb data files equal to your CPU count up to 8 data files (Don't create more than 8 data files). So, if your CPU count is 4, then create 4 identically sized tempdb data files. If you CPU count is 16, then only create at most 8 data files.
Frankly I prefer using PURE storage until management got convinced into moving to the cloud. 10x faster than any cloud storage and has advanced deduplication technology so a 1TB database would actually only use 100GB of storage. Disk latency was always less than 1ms using PURE storage. Now I'm getting 20 -50 ms latency using azure storage or aws ec2.
Tung Dang
Azure and SQL Server DBA Contractor / Consultant
SQL Brainbox - SQL Server Monitoring Tool
Viewing 7 posts - 1 through 7 (of 7 total)
You must be logged in to reply to this topic. Login to reply