pdpv0: add a simple status check task

[beck-8]

add some simple status checks for /pdp/ping

• db check
• height diff check
• other ?

Will doing some simple, time-consuming checks help troubleshoot PDP node problems faster?
can also choose a better node for Synapse SDK

[LexLuthr]

The response below is more general in nature and is intended for a wider audience rather than being specific to this issue.

This is becoming a problem. Soon, we will need to expose the whole DB to the world for debugging. What happens when all these checks pass and still you fail to onboard the data? Do we keep adding stuff or just take over the SP at some point?

We are trying to guarantee the QoS when we don't control the system. How does that make sense? If we really want a proper QoS based setup then a middle man has to sit in between SDK and Curio. This service would be responsible for health check and QoS maintenance and SP selection. Curio can expose some dedicated endpoints to this service without having to worry about hundreds of health check request from random IPs.

QoS guarantees must be enforced from protocol level and SP operation is the only one capable of ensuring it. If there is any expectation that Curio will somehow magically fix issues with administration then we have a serious goal mismatch.

PDPv1 already have solid logging and easier GUI/Log based debugging and a proper pipeline structure with a single endpoint to tell you what's the status of onboarding. But, we keep building patches on top of already patchy implementation.

[ZenGround0]

We are going to add a HealthTask which will print this out periodically so that betterstack logging can capture this but we are not exposing it to the world.