I'm looking for any specific architecture information with regard to how the APEX ORDS middle-tier layer scales on an ADB implementation. Specifically, as a factor of changing the OCPUs for the ADB, does that APEX ORDS middle-tier also scale as well? Thanks for any insight.
Oracle ORDS middle-tier scalability, as it ties to OCPU allocation on Oracle Autonomous DB
RayC1
0
Viewed 24
1 Answers
There is no relation to the number of oCPUs on your Autonomous Database and the Oracle REST Data Services mid-tier deployments.
Each region has X ORDS nodes deployed, front-ended by a LoadBalancer.
Each of those X nodes will have a connection pool for each PDB, 100 connections each.
As demand increases in a region, the managed ORDS environment can quickly be increased by our DevOps team (either by button press or automation).
You can handle quite a few concurrent user requests if you have 300, 400, 500, 600, 700, even 1200 connections available.
If not, then you might consider:
- making sure your SQL and PL/SQL are optimal for HTTPS, i.e. making it FAST
- deploy a customer-managed instance of ORDS on your own tenancy compute node - you setup ORDS and give it as many resources as you want. Of course at this point, you're managing your mid-tier, not us.
Related