> Originally this was caused by an Istio sidecar pod reaching its concurrency limits and failing to auto scale correctly because of a misconfigured policy that watched host service but not sidecar limits.
I operated services at similar scale, and generally we use to put a bit of slack so that you would get an alarm when capacity goes up to 80%+ (or whatever number makes sense)
This allows to check, in the morning, after coffee, why the load balancer fleet didn't scale up automatically.
I am sure there is a good answer to why this is impractical, but it would be nice to know
I'm not saying they're not useful, of course they are. I am disputing that they are part of the agreed bargain between you and the proprietary LLM providers.
They explicitly do not promise reasoning traces. You (general you) agree to those terms and pay for that bargain anyways.
I’m just driving by here but they bill by tokens — it’s a stretch to turn around and deny your right to see them. And it’s especially egregious when the tokens admittedly, routinely do the opposite of what you instructed.
But personally it’s not about right and won’t it’s just blatant bullshit.
And lawyers bill by 6-minute increments, yet that doesn't mean you get access to all of a law firm's internal discussions and notes about you and your case.
Just because you paid for the lawyer time/LLM tokens doesn't mean you get access to everything that happened within that time/tokens.
Physics. Right now the clothes folding is computationally limited, but assuming we get faster computers to get them folding faster, gravity and air resistance are still going to take their time.
I am afraid that this post is missing the biggest point.
Given a prompt (or task) how do I evaluate if it is a "simple" task that should be executed by a small model or if it is a complex one that may need a SOTA model?
Feel free to take a look at the docs for how routing decisions are made: https://role-model.dev/
When you use Pi with pi-role-model, Pi will include task and role metadata with its request; the role-model router runtime additionally holds benchmark and observability data, and a configured routing strategy. A composite of this is used to make the decision.
What you are pointing out is correct: making the actual decision and ensuring it is accurate can be difficult, which is why you need rich data as above, and also a model pool where each model is distinct, as I wrote in this post and in one of the comments below.
I operated services at similar scale, and generally we use to put a bit of slack so that you would get an alarm when capacity goes up to 80%+ (or whatever number makes sense)
This allows to check, in the morning, after coffee, why the load balancer fleet didn't scale up automatically.
I am sure there is a good answer to why this is impractical, but it would be nice to know