3 ms·
Would inclusion of a small random timeout before serving the request solve this? Would not the load balancer activity cancel the effect? If anything, adding ra
by nine_k 10y ago
Would inclusion of a small random timeout before serving the request solve this? Would not the load balancer activity cancel the effect?
If anything, adding random garbage to the error page to approximately match the size of a normal response should make the attack useless.
- pfg 10y agoTypical (WAN) network conditions are not all that different from a small random timeout, and these kind of attacks still work. Adding random noise just means you'll need more samples, but statistics are going to win eventually. I guess a combination of random noise and tight request rate limiting might mitigate this in practice (in exchange for a DoS vector, I suppose).