Running SQRL as a Server

Starting

One way of running SQRL in production is by running it as a stateless server:

1
2
3
4
5
6
7
8
9
10
$ cat << EOF > ratelimit.sqrl
LET Ip := input();
LET Remaining := rateLimit(BY Ip MAX 2 EVERY 30 SECONDS);
CREATE RULE BlockedByRateLimit WHERE Remaining = 0
  WITH REASON "Saw more than two requests in the last thirty seconds.";
WHEN BlockedByRateLimit THEN blockAction();
EOF

$ ./sqrl serve --port=2288 ratelimit.sqrl
Serving examples/ratelimit.sqrl on port 2288

Once your server is up and running, it will serve traffic over HTTP. You can test it out with the curl command line tool, sending it a couple sample request from the IP address 1.2.3.4:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
$ curl -d '{"Ip": "1.2.3.4"}' \
    -H "Content-Type: application/json" \
    'localhost:2288/run?features=BlockedByRateLimit,Remaining&pretty'
{
  "allow": true,
  "verdict": {
    "blockRules": [],
    "whitelistRules": []
  },
  "rules": {},
  "features": {
    "BlockedByRateLimit": false,
    "Remaining": 2
  }
}

# If you send enough requests in a short enough time
$ curl -d '{"Ip": "1.2.3.4"}' \
    -H "Content-Type: application/json" \
    'localhost:2288/run?features=BlockedByRateLimit,Remaining&pretty'
{
  "allow": false,
  "verdict": {
    "blockRules": [
      "BlockedByRateLimit"
    ],
    "whitelistRules": []
  },
  "rules": {
    "BlockedByRateLimit": {
      "reason": "Saw more than two requests in the last thirty seconds."
    }
  },
  "features": {
    "BlockedByRateLimit": true,
    "Remaining": 0
  }
}

State storage and scalability

By default the command line tool stores all state in memory. Enabling Redis (or any external database) state storage will give you a stateless server that you can scale horizontally.

Sync and Async Paths

We recommend running every event through two separate SQRL configurations. Read our Async deployment documentation for more information.