5 ms·

6M per day is an average of 70rps. It would be highly unusual to have a 100x variation from the mean on such a large traffic volume. Let's imagine that all the

6M per day is an average of 70rps. It would be highly unusual to have a 100x variation from the mean on such a large traffic volume. Let's imagine that all the traffic occurs in a single hour (6M per hour) and is Poisson distributed. Even then to reach 10,000 rps would require a second with 6x the average traffic volume for that hour. Admittedly this could theoretically occur when some correlated event causes everyone to browse HN, but historically the site goes down during those. For comparison our service (also largely US-centric) served 30M views yesterday with a peak volume of ~1000rps.

sulam 5y ago
I'm going to assume that since you are referring to exponential distributions in one part of your response that you're familiar with this topic, even though you refer to averages in another part. Averages are, as you likely know, completely irrelevant when it comes to predicting peak load. Yes, 10K is likely an overestimate by some factor, however it's in the realm of possible. We can argue how much of an overestimate, but I don't think this takes away from the main point that I was trying to make above, which is that they are very unlikely to exceed the serving capacity of a single host (I don't know how fast Arc is, but a well-tuned serving system should be able to handle 40K rps for simple to cache content, which HN surely qualifies as).
- ericpauley 5y ago
  Under the Poisson process assumption the average rate directly characterizes the distrubtion, so it is very relevant to this discussion. The non-Poisson component of the phenomenon (correlated spikes in traffic due to events) are not well-characterized via averages, which is why the original analysis relaxed that limitation by examining a single worst-case traffic hour. Beyond that the Poisson process analysis is standard. While 10k rps may be possible it would either require vastly exceeding 6M requests per day (and per hour) or some very non-standard traffic event (this could happen in practice, for instance email asset delivery is increasingly peaky due to modern email clients and can exhibit very large traffic spikes). As HN is primarily visited by human users spanning many organizations such a strongly correlated event seems unlikely.
  - sulam 5y ago
    Again, you seem to be nitpicking my upper bounds estimate of their traffic. Fair enough. Are you making any broader point other than that we may mildly disagree on what that bound is? If so, that's fine, I picked a higher round number than is likely. This simply reinforces the broader point I was making.

sulam 5y ago

I'm going to assume that since you are referring to exponential distributions in one part of your response that you're familiar with this topic, even though you refer to averages in another part. Averages are, as you likely know, completely irrelevant when it comes to predicting peak load. Yes, 10K is likely an overestimate by some factor, however it's in the realm of possible. We can argue how much of an overestimate, but I don't think this takes away from the main point that I was trying to make above, which is that they are very unlikely to exceed the serving capacity of a single host (I don't know how fast Arc is, but a well-tuned serving system should be able to handle 40K rps for simple to cache content, which HN surely qualifies as).

ericpauley 5y ago
Under the Poisson process assumption the average rate directly characterizes the distrubtion, so it is very relevant to this discussion. The non-Poisson component of the phenomenon (correlated spikes in traffic due to events) are not well-characterized via averages, which is why the original analysis relaxed that limitation by examining a single worst-case traffic hour. Beyond that the Poisson process analysis is standard. While 10k rps may be possible it would either require vastly exceeding 6M requests per day (and per hour) or some very non-standard traffic event (this could happen in practice, for instance email asset delivery is increasingly peaky due to modern email clients and can exhibit very large traffic spikes). As HN is primarily visited by human users spanning many organizations such a strongly correlated event seems unlikely.
- sulam 5y ago
  Again, you seem to be nitpicking my upper bounds estimate of their traffic. Fair enough. Are you making any broader point other than that we may mildly disagree on what that bound is? If so, that's fine, I picked a higher round number than is likely. This simply reinforces the broader point I was making.