4 ms·

Isn't there the assumption in this article that tasks will come in faster than one person can deal with them, which of course leads to an exponential/snowball e

Isn't there the assumption in this article that tasks will come in faster than one person can deal with them, which of course leads to an exponential/snowball effect. To then add a second person, keeping the same input, isn't it obvious that the task latency will drop more than linearly? (Because two ppl working on tasks coming in will be able to respond in the minimal time, rather than them banking up) Task management, nor stats, are part of my wheelhouse, so I assume I missed some glaring detail about this article?

ywei3410 6y ago
> Isn't there the assumption in this article that tasks will come in faster than one person can deal with them Yes this is an assumption. This isn't important though - you use a queue when you think tasks will have some level of build-up anyway i.e. the chance of the task being above your processing speed is non-zero. > Because two ppl working on tasks coming in will be able to respond in the minimal time, rather than them banking up No - that's not what the article is suggesting at all. It's saying that when you have two workers, the relative probability that both workers are queued up at the same time, causes the loss of the quadratic relationship between latency and N (see formula at the end). EDIT: Clarification on the first part. It is not that they will come in faster than the processing speed, but they that they can.
rmrfstar 6y ago
I grepped the article for "Amdahl" before reading, didn't see it and stopped reading. It's my version of not reading specifications for perpetual motion machines.
- fis 6y ago
  This article isn't about Amdahl's law, it's about queueing theory. Little's law would be more applicable.
  - rmrfstar 6y ago
    Didn't even read the article, just the title.
btilly 6y ago
Isn't there the assumption in this article that tasks will come in faster than one person can deal with them, which of course leads to an exponential/snowball effect. They are coming in at exactly the speed that one person can deal with them. Which means that in the long run the worker is always working, and lines can get arbitrarily long. That is what causes his "quadratic". A second worker makes the odds of a long line drop dramatically.
- baking 6y ago
  I may be reading it wrong, but if there is a 50% chance that a worker can finish a task in one turn, two workers would help keep the backlog down to linear time. But I learned it from Alvin Drake in 6.041.
learnstats2 6y ago
Is adding a second worker just equivalent to halving the amount of work? I don't think this article answered that question. Since the problem is really that the first worker is overburdened, by being asked to work at maximum capacity plus randomness.
eru 6y ago
> Isn't there the assumption in this article that tasks will come in faster than one person can deal with them, which of course leads to an exponential/snowball effect. Not everything that grows, not ever everything that grows fast, grows exponentially.

ywei3410 6y ago

> Isn't there the assumption in this article that tasks will come in faster than one person can deal with them Yes this is an assumption. This isn't important though - you use a queue when you think tasks will have some level of build-up anyway i.e. the chance of the task being above your processing speed is non-zero. > Because two ppl working on tasks coming in will be able to respond in the minimal time, rather than them banking up No - that's not what the article is suggesting at all. It's saying that when you have two workers, the relative probability that both workers are queued up at the same time, causes the loss of the quadratic relationship between latency and N (see formula at the end). EDIT: Clarification on the first part. It is not that they will come in faster than the processing speed, but they that they can.

rmrfstar 6y ago

I grepped the article for "Amdahl" before reading, didn't see it and stopped reading. It's my version of not reading specifications for perpetual motion machines.

fis 6y ago
This article isn't about Amdahl's law, it's about queueing theory. Little's law would be more applicable.
- rmrfstar 6y ago
  Didn't even read the article, just the title.

btilly 6y ago

Isn't there the assumption in this article that tasks will come in faster than one person can deal with them, which of course leads to an exponential/snowball effect. They are coming in at exactly the speed that one person can deal with them. Which means that in the long run the worker is always working, and lines can get arbitrarily long. That is what causes his "quadratic". A second worker makes the odds of a long line drop dramatically.

baking 6y ago
I may be reading it wrong, but if there is a 50% chance that a worker can finish a task in one turn, two workers would help keep the backlog down to linear time. But I learned it from Alvin Drake in 6.041.

learnstats2 6y ago

Is adding a second worker just equivalent to halving the amount of work? I don't think this article answered that question. Since the problem is really that the first worker is overburdened, by being asked to work at maximum capacity plus randomness.

eru 6y ago

> Isn't there the assumption in this article that tasks will come in faster than one person can deal with them, which of course leads to an exponential/snowball effect. Not everything that grows, not ever everything that grows fast, grows exponentially.