A rateLimit rule counts the requests that match its conditions, per client IP. A client over the
limit is blocked (403) until its count in the window falls back under the limit. Requests under the
limit carry on to the next rule.
{
"name": "login",
"action": "rateLimit",
"mode": "block",
"rateLimit": { "limit": 20, "windowSeconds": 300 },
"conditions": [
{ "type": "equals", "negate": false, "target": "path", "value": "/login", "caseSensitive": false },
{ "type": "equals", "negate": false, "target": "method", "value": "POST", "caseSensitive": false }
]
}| Field | |
|---|---|
rateLimit.limit |
1 to 100,000 requests |
rateLimit.windowSeconds |
60, 300 or 600 |
mode |
block (the default), or log to record clients over the limit without blocking them |
- To count every request, use a condition that always matches, such as
pathregex^/. - Counting is per client IP. Clients behind one address, such as an office’s network, share a count.
- A rate limit can’t have a
bodycondition, or come after a rule that has one.
Set a new limit’s mode to log first, check in the logs which clients go over, then
change it to block. See Rate limit a login page.