Vivekanandha Institutions — Platform Architecture

Infrastructure diagram

Click a node or connection for details. Scroll to zoom, drag the background to pan. Hover a node to trace its connections. The overlay colours components by the simulator's load at the selected time.

Load simulator

Scenario Peak concurrent users Database tier Cache hit ratio Neo4j plan Neo4j writes User lookups

Concurrent users & error rate

p95 latency by workload (ms, log scale)

Utilisation of the busiest components

Components at the selected time

Capacity arithmetic at peak (steady state, fully scaled)

Workload assumptions (per-user request rates and per-request costs)

Monthly cost

Follows the simulator: scenario, users, database tier, cache and Neo4j plan. Estimates — see sources below.

Teaching hours per day Peak hours per day Average load (share of peak) Night load (share of peak)

Line items

New app impact check

Describe an app you want to add. See what it does to the database and every existing app before you build it.

The app Name What it does Concurrent users at its busiest Usage profile Requests per user per minute

How it is built Runs on Workers per instance Max instances Mongo pool size per worker Uses blocking libraries in async code (sync PyMongo, boto3, smtplib…)

Data Database Dedicated cluster tier Users, roles and colleges from Writes into shared collections (users, user_groups…) Uses Neo4j (graph) Graph operations per request

Busy during these scenarios: