--server-max-concurrency
- CLI
- --server-max-concurrency
- REST
- CLI only
Caps the renders running at once, at least 1. A request past the cap waits in the --server-queue-requests queue.
The cap is 1 without an Enterprise license, and defaults to the CPU count with one.
Sizing
Each running render is a full browser. Plan memory per slot, on top of the server’s own:
| Render | Memory |
|---|---|
| Video | Up to 1 GiB per slot |
| Crawl or web archive | About 500 MiB per slot |
| Full-page screenshot | 200 to 300 MiB per slot |
| Single-page PDF, Markdown or HTML | Under 200 MiB per slot |
| The server | About 100 MiB, once |
Four concurrent video renders peak near 4 GiB, so give them a 6 GiB host. Memory varies with the page rendered, so measure your own workload before setting the cap.
Example
CLI
Run locally
zshot --server --server-max-concurrency 4