Rewrite submit_block relay communication to support SSZ and JSON#468
Rewrite submit_block relay communication to support SSZ and JSON#468JasonVranek wants to merge 1 commit into
Conversation
JasonVranek
force-pushed
the
pr/ssz-routes
branch
2 times, most recently
from
c800921 to
cbf451d
Compare
JasonVranek
force-pushed
the
pr/ssz-routes
branch
from
a52b531 to
2565c31
Compare
JasonVranek
force-pushed
the
pr/ssz-submit-block
branch
from
d468365 to
3740824
Compare
content negotiation. Includes: - SSZ-first request encoding with JSON fallback on 406/415 - Content-Type and Eth-Consensus-Version header handling - Fork-aware SSZ decoding for relay responses - MIME parameter tolerance on relay response Content-Type - v2 to v1 fallback forwards payload to BN (prevents silent block loss) - V2 fallback metric counter - Comprehensive submit_block integration tests for both encodings
JasonVranek
force-pushed
the
pr/ssz-submit-block
branch
from
3740824 to
0d76e5c
Compare
| Ok(response) => { | ||
| // If the original request was for v2 but we had to fall back to v1, return a v2 | ||
| // response | ||
| // If the original request was for v2 but we had to fall back to v1, the |
There was a problem hiding this comment.
Would be great if these fallbacks could be removed. The v1 endpoint has been deprecated for a long time now. But relays still support it because boost clients still call it and boost clients still call it because it's available as a fallback.
I'd be surprised if there are still relevant relays that don't yet support v2 at this point.
There was a problem hiding this comment.
agreed it would be nice to prune the fallbacks but in the off chance the edge case is hit, this is purely more redundancy with no latency penalty. (I believe during kurtosis testing we hit this all the time too hence why it was added). I'd argue keep for now and after gloas v1/v2 are deprecated anyways
There was a problem hiding this comment.
Yeah fair point about Gloas. Hardly worth trying to remove it now at this stage when gloas is not too far away.
| .post(url.as_ref().clone()) | ||
| .timeout(Duration::from_millis(timeout_ms)) | ||
| .headers(headers) | ||
| .body(serde_json::to_vec(&signed_blinded_block).unwrap()) |
There was a problem hiding this comment.
unwrap -> map_err into PbsError ?
| res = match relay | ||
| .client | ||
| .post(url.as_ref().clone()) | ||
| .timeout(Duration::from_millis(timeout_ms)) |
There was a problem hiding this comment.
because this doesn't run concurrently with the previous call, this function can take timeout_ms x2 and then the submit_block_with_timeout's accounting of remaining_timeout_ms is incorrect:
remaining_timeout_ms = timeout_ms.saturating_sub(start_request.elapsed().as_millis() as u64);| ) -> Result<SubmitBlindedBlockResponse, PbsError> { | ||
| let data = PayloadAndBlobs::from_ssz_bytes_by_fork(response_bytes, fork).map_err(|e| { | ||
| PbsError::RelayResponse { | ||
| error_msg: (format!("error decoding relay payload: {e:?}")).to_string(), |
| .timeout(Duration::from_millis(timeout_ms)) | ||
| .headers(headers) | ||
| .body(serde_json::to_vec(&signed_blinded_block).unwrap()) | ||
| .header(CONTENT_TYPE, EncodingType::Json.to_string()) |
There was a problem hiding this comment.
EncodingType::Json.content_type_header()
| .json(&signed_blinded_block) | ||
| .headers(headers.clone()) | ||
| .body(signed_blinded_block.as_ssz_bytes()) | ||
| .header(CONTENT_TYPE, EncodingType::Ssz.to_string()) |
There was a problem hiding this comment.
EncodingType::Ssz.content_type_header()
| relay, | ||
| url, | ||
| timeout_ms, | ||
| proposal_info.headers.clone(), |
There was a problem hiding this comment.
let's pass by reference, we need to clone (not move) inside once anyway
3 participants
Summary
Rewrites submit_block relay communication to support SSZ and JSON content negotiation.
What's in this PR
V2_FALLBACK_TO_V1metric counter for observabilityReview notes
The v2→v1 fallback behavior is an intentional change from the previous code which silently discarded the v1 payload. The beacon node needs the unblinded execution payload to broadcast the block (returning an empty 202 on fallback would cause silent block loss).
Part 3 of 4: SSZ types → get_header → submit_block → validation bypassing
Attribution
This work builds on the SSZ builder flow originally implemented by @eserilev
in #252 and ported to the current codebase by @jclapis in #403.
Co-authored-by: Eitan Seri-Levi eserilev@ucsc.edu
Co-authored-by: Joe Clapis jclapis@outlook.com