Retry PSD with valid deals if the message failed

[jennijuju]

Checklist

• This is not a new feature or an enhancement to the Filecoin protocol. If it is, please open an FIP issue.
• This is not a new feature request. If it is, please file a feature request instead.
• This is not brainstorming ideas. If you have an idea you'd like to discuss, please open a new discussion on the lotus forum and select the category as Ideas.
• I have a specific, actionable, and well motivated improvement to propose.

Lotus component

• lotus daemon - chain sync
• lotus miner - mining and block production
• lotus miner/worker - sealing
• lotus miner - proving(WindowPoSt)
• lotus miner/market - storage deal
• lotus miner/market - retrieval deal
• lotus miner/market - data transfer
• lotus client
• lotus JSON-RPC API
• lotus message management (mpool)
• Other

Improvement Suggestion

Currently, if a PSD is failed, SP cannot re-publish the valid deals in a new msg - however, we should do that to improve deal-making experience for both sides.

[nonsense]

@jennijuju could you clarify this ticket? Is this about retrying deals that were batched together with a failing deal?

AFAICT when we have 1 deal in a PSD message, and it fails, it is not a transient error, and there really is nothing to retry. If we have N deals and a PSD message fails because of 1 bad deal, then it makes sense to still retry the N-1 deals, but this obviously is not how Lotus works today, and something worth improving. WDYT?

[jennijuju]

@jennijuju could you clarify this ticket? Is this about retrying deals that were batched together with a failing deal?

AFAICT when we have 1 deal in a PSD message, and it fails, it is not a transient error, and there really is nothing to retry. If we have N deals and a PSD message fails because of 1 bad deal, then it makes sense to still retry the N-1 deals, but this obviously is not how Lotus works today, and something worth improving. WDYT?

This is exactly for the later case!

[nonsense]

@jennijuju sorry, but having looked at the code, this is still confusing.

We already validate and publish only deals that are valid, done at #7234 . So I assume this ticket is about retrying PSD after it has actually failed, which means that it passed simulation, only valid deals were included because of simulation, but even so - the PSD failed for some reason.

I am trying to understand why a PSD would fail after simulation, before trying to address this.

[nonsense]

Having spoken to Lotus folks, we decided to close this one in favour of #7102

Basically it is not obvious when an error is transient and when we need to retry, so for the time being having a manual retry mechanism is a good first step.