If you don't need (micro)services, don't build them -- they only introduce complexity in communication between components, build tooling, diagnosing issues, and they slow things down. It's amazing what you can do on a single machine, or, beyond that scale, a simple two-tier architecture with a few web nodes and a big SQL database. Systems like that are much cheaper, more predictable, and easier to understand. Of course, if you're Google or Facebook scale, distributed systems are (a) necessary (evil).
Additionally, I'm not sure his last point "extract services" is a good rule of thumb. If something can be a library, it'll be much simpler as a library. Don't make an "email service" when you can just import an SMTP library, wrap it in a couple dozen lines of code, and be done with it.
2016 https://news.ycombinator.com/item?id=12245909
Discussed at the time: https://news.ycombinator.com/item?id=5055371
I was never great at stat's. Can somebody explain why this is?
What has the average told us? It has told us nothing because it's meaningless with a bi-modal distribution.
For more info read up on the normal distribution, guassian distribution, bi-modal distributions, and the central limit theorem.
The average latency does not tell us much since it does not really represent anything meaningful - i.e. it is not the latency of typical user/request as you might expect since the average is being skewed by the extremely high tail-latencies [2] or by the multiple modes (the peaks in latency histogram). The typical latency would be more likely better represented by median latency (not in the multimodal distribution case afaik).
As for why not go just go with median latency: you usually need to make multiple requests in parallel and end up waiting for the slowest request of the group. The 95th, 98th or 99th percentile is commonly used to cover for this (sorry can't find a suitable reference). This is also preferable in case of the multimodal distribution (well at least for monitoring and/or general performance diagnosis purposes since you usually care about the worst/tail cases).
[1]: https://en.wikipedia.org/wiki/Multimodal_distribution
[2]: https://en.wikipedia.org/wiki/Skewness#/media/File:Relations...
I think it is because of outliers and how diatributions can look like, if you only know the mean.
1. A single extreme outlier can move your mean far away from where you would expect it to be, seeing all the other values.
2. If you see the mean alone, you might thing values are around that value, but not even a single value has to be anywhere cloe to the mean. For example: 1 1 1 1 9 9 9 9 → means is 5. Pathological example, but you can mix the numbers however you want and there are infinitely more examples, where the mean could mislead you.
3. The mean is part of the arguments to the formula of a Gaussian (bell shaped curve, normal distribution). The other is the standard deviation. So if you know the mean, you have already some valuable info on the bell shaped curve. You already know where its center is. That is, why giving the mean makes sense in that case.
Surely others can put it better or add lots of reasons.
