Docs

Install, load your data, query it, and reproduce every number on this site.

1. Install

sudo ./semurg-install.sh

Override anything the hardware scan picked:

sudo SEMURG_STRIPE_ROOTS=/data0/semurg,/data1/semurg \
 SEMURG_PORT=8080 ./semurg-install.sh

2. Check the instance

curl http://localhost:4000/v1/status

Reports the licence, whether the instance is writable, and where the customer-data store lives. Writes require a licence carrying the write capability; without it the instance is read-only and says so.

3. Load your data

Batch-first by design. One request carries a batch of 64-byte containers and costs exactly one crossing into the engine. There is no per-item endpoint, because a per-item turnstile is the one shape that cannot reach the engine's throughput. Up to 65,536 containers per request; send successive batches for more.

curl -X POST http://localhost:4000/v1/ingest \
  -H 'content-type: application/json' \
  -d '{"containers_b64":"<base64 of packed 64-byte containers>"}'

{"ok":true,"containers":65536,"bytes":4194304,
 "elapsed_us":6647,"containers_per_s":9859485}

4. Query it

curl -X POST http://localhost:4000/v1/query \
  -H 'content-type: application/json' \
  -d '{"op":"fetch","ids":[1,2,3]}'

op	takes	does
fetch	ids	batch point read
resolve	id	id to physical location
search	needle	scan-at-bandwidth pattern match
histogram	field_offset, buckets	group-by aggregation

5. Reproduce our numbers

curl http://localhost:4000/api/dashboard

The same call works on this site and on your instance. It names the box that produced the numbers, so you can put them side by side. That is the point: do not take our word for it, run it on your hardware.

Licence

Verification is fully offline, Ed25519, no network call ever. The engine refuses to open a store without a valid unexpired licence. Expired, tampered and missing keys all fail closed with a clear message. Air-gapped deployments never phone home.

Managing it

systemctl status semurg
journalctl -u semurg -f
systemctl restart semurg

An honest note on claims

We do not claim a zero-index database. Truth is the value log; indexes are bounded and re-derivable from it. The bet is bandwidth: scan and fuse at the drive's sequential ceiling instead of maintaining a pile of structures that each need their own copy of your data. Where that bet loses, we say so on the pricing page.