Redis Connection Pool Calculator

Give the instance count and pool size. The result shows the share of maxclients used, the buffer memory it costs, and what a single connection can actually sustain.

Paste below, or drop a file anywhere on this panel

Or drop a file anywhere on this panel. Nothing is uploaded: the analysis runs in this tab.

The answer appears here

Paste on the left and press Calculate. Nothing leaves this tab.

Examples

Real input you can load into the tool above. Each one shows a different thing going wrong, because that is what the tool is for.

An oversized pool

Enough connections to approach maxclients, which refuses new ones at the limit

app-instances: 500
pool-size: 20
maxclients: 10000

A reasonable pool

Fewer connections, with the throughput a single one sustains

app-instances: 20
pool-size: 10
maxclients: 10000

Common mistakes

These are the ones that fail silently. The config is accepted, nothing raises an error, and the consequence arrives later.

  1. Raising the pool size to increase throughput

    Redis executes on one thread, so all connections share one executor. Adding connections adds contention and buffers, not capacity.

    Instead:Pipeline instead. That genuinely raises throughput.

  2. Ignoring the memory each connection costs

    Every connection carries input and output buffers, roughly 20 KB idle and far more under load.

    Instead:Include it. Ten thousand connections is hundreds of megabytes before any data.

  3. Forgetting a deploy can double the count

    During a rolling deploy both old and new instances hold pools, briefly doubling connections.

    Instead:Leave headroom below maxclients for the overlap.