Read+Write
TPS
1400 K
Latency
11 ms p(99)
Cost
$11.77
per month / per 1K TPS
Read Only
TPS
1358 K
Latency
10 ms p(99)
Write Only
TPS
42 K
Latency
20 ms p(99)
Configuration
Read / Write Allocation: 97% / 3%
Data Unit Size: 500 bytes
Testbed: 4 Gatling instances, 16 cores, 10 minutes of max throughput
App: Akka 2023.10, 40K entities, event sourced log, source code
Infrastructure:
1 GKE cluster with 36 n2-std-8 nodes ($10,996.09 / mo with 1-yr full upfront)
16 PostgresSQL with 8 vCPUs 32GB memory, 250 GB SSD ($343.12 / mo with 1-yr full upfront)
Total Monthly Cost: $16,485.93
Definitions
TPS: # of data units processed (read & write) per second
Latency: round trip HTTP request-response time measured during peak TPS
Cost: Total reserved instance costs for a 31 day month ratioed against 1K TPS
This was a sample test to demonstrate that 1,400,000 req/s of load with 97% reads and 3% writes was possible with Akka whose events are persisted to a Postgres database. This application simulates a digital twin solution, where a single Akka actor represented an IOT device that had to persist 500 bytes of information on each update.