Fine tuning CPU

Hi <@489175443153354772> I was running Perf testing using rust based subgraphs.
Here are the numbers I am seeing for 20kb payload size and 100 kb payload size with different core.
Even though I increase the core I do not see increase in TPS for lower payload. Between 3 and 4 core for 20 kb payload I do not see any difference at all.
For larger payload, I did see increase in throughout between 2 and 3 core however again no difference between 3 and 4 core .

Have you tested against the subgraphs individually without the gateway?
Can you provide a repository to replicate your findings?

This is without gateway

When I ran the tests my subgraph was set to 2 core and min and max pods were 10-20. However gateway scaling was disabled.

What is the max number of requests wundergraph can serve at any given time? Can we configure that ?

I suggest that you share the repository you’re using to benchmark. Otherwise, it might be hard to replicate or fully understand your findings.