Hello OMOP community.
I was wondering if anybody has any best practices or recommendations for a typical storage subsystem to support a typical OMOP database environment. What type of IOPS and throughput would be sufficient?
Thank you
Boris
Hi @bbelous (and @jchang)! Welcome to the OHDSI Community. When you have a moment, consider Introducing Yourself to the community so folks know more about you!
I know you’ve posted a couple times in the last month asking this type of question. Sorry these were missed. August is peak holiday time for many of our European collaborators. We also were all pretty consumed last week watching @Patrick_Ryan @aostropolets and team challenge us to reproduce a study! If you didn’t tune in, definitely recommend looking at the content from 2021 Global Symposium.
On the best practices side, there’s a few places to start!
- 2021 OHDSI Symposium Collaborators Showcase
- A few humans: @Ajit_Londhe @Chris_Knoll @gregk @mvanzandt @anthonysena @aandryc @esholle et al might have some places to direct you on infrastructure efforts or posters/talks they’ve presented in the past (each are masters in their ETL understanding and have their own environments in tow). There was an infrastructure WG a while back but it’s been some time. Not sure if it’s active.
Hope this helps!
Best,
Kristin