| CHIP Number | 0033 |
|---|---|
| Title | Add additional partial headers |
| Description | Add farmer and harvester peer id and version as well as harvester raw/effective capacties and plot count to partial headers |
| Author | Felix Brucker |
| Editor | Dan Perry |
| Comments-URI | CHIPs repo, PR #114 |
| Status | Final |
| Category | Informational |
| Sub-Category | Informative |
| Created | 2024-03-29 |
| Requires | None |
| Replaces | None |
| Superseded-By | None |
Add farmer and harvester peer id and version as well as harvester raw and effective capacities and plot count to the header of partial submit requests to pools, so that pools can correctly show farmers with their version and harvesters with their respective version, capacities and plot count.
Currently, a partial submitted to a pool contains the following information (among others):
- the harvester peer id of the harvester who found the proof of space as part of the partial payload
- the farmer version as part of the
user-agentheader
This does not allow pools to correctly map a version to a harvester (as they only know the farmers version) and it does not allow them to map an identity to the farmer version. This means currently pools can not correctly show the version of their users' software.
For example: A pool knows that harvester A is using a farmer with version x.y.z, but it does not know the version of harvester A. In addition it can not create a farmer entity to attach additional pool specific functionality (like a name, version and vendor, notification settings, connected harvesters, total space, etc.) to, as there is no id to link the x.y.z farmer version to.
The following is an example of how a harvester overview in a pool looks like, when it includes the additional information. A farmer overview could look similar.
Implementing this change is very easy and has already been done, see below.
This proposal only adds http request headers and is fully backwards compatible to the current implementation.
Adding the information to http headers allows for concise naming and easy retrieval of values without complex extraction (for example extracting the farmer version from the user agent header) without modifying the partial payload, which could cause problems.
Introduce two new config fields:
farmer.include_harvester_stats_in_partials: boolean which defaults totrue. This option determines whether harvester capacities and plot counts are included in partial submit headers.farmer.include_peer_ids_in_partials: boolean which defaults totrue. This option determines whether the farmer and harvester peer ids are included in partial submit headers.
The following new headers will be included in partial submits:
chia-farmer-version: the full farmer version, same as already part of the user agent, for example:2.2.1.giga35chia-harvester-version: the full harvester version, for example:2.2.1-dr-0.15.0
The following headers are added when farmer.include_peer_ids_in_partials is true:
chia-farmer-peer-id: the farmer peer id (bytes32 as hex), for example:ca93b47f157b4be9e91e88943a3c7ab9c2d6def0d7bd42dd0d468e54a14812f6chia-harvester-peer-id: the harvester peer id (bytes32 as hex), already part of the partial payload, included for consistency, for example:f7959d25848dcccbf684f44951272d0a2323e1511f90e3f44651e1e10463eeb9
As well as the following optional headers, based on availability and whether the farmer.include_harvester_stats_in_partials is true:
chia-harvester-raw-capacity-bytes: the total raw capacity in bytes of this harvester and PlotNFT, for example:1099511627776chia-harvester-effective-capacity-bytes: the total effective capacity in bytes of this harvester and PlotNFT, for example:4398046511104chia-harvester-plot-count: the total plot count of this harvester and PlotNFT, for example:1337
The current implementation in the form of a PR is available here
This proposal only adds additional headers to the https secured pool partial requests.
Copyright and related rights waived via CC0.