perf(router): skip upstream gzip auto-decode on smart-router HTTP req…

[NadavLevi]

…uests

Go's default http.Transport auto-adds Accept-Encoding: gzip on outbound requests when no Accept-Encoding is set, and transparently decodes responses via net/http.(*http2gzipReader).Read → compress/gzip.Reader → compress/flate. Production pprof on the eth router attributed ~30-39% of total CPU to that inflate chain — CPU we spend decoding bytes we then either pass through, re-encode with gzip for the client, cache as a blob, or peek at with CheckResponseError.

Two coordinated changes remove that work only on the smart router:

1. Switch lavasession.HTTPDirectRPCConnection from its own fresh &http.Client{} (default transport — no pooling, no TLS session cache, no HTTP/2) to common.OptimizedHttpClient(). Every smart-router HTTP connection now shares common.SharedHttpTransport() and inherits connection pooling, TLS session resumption, and ForceAttemptHTTP2.

2. Set Accept-Encoding: identity on every outbound request from HTTPDirectRPCConnection.SendRequest and DoHTTPRequest. Go only auto-adds gzip when the caller left Accept-Encoding empty; once identity is set, both the auto-add and auto-decode are skipped.

The scoping matters: the shared transport keeps its default compression behavior, so provider chain proxies (chainlib/rest.go, tendermintRPC.go, chainproxy/rpcclient/http.go) continue to auto-gzip as before. The opt-out lives exclusively on the smart-router request path.

Measured impact on the eth router (same 900s pprof window, back-to-back before/after the deploy):

CPU cores (avg): 2.47 → 1.23 (-50%)
Memory inuse: 277 GB → 186 GB (-33%)
Gzip-decode CPU path: 38.9% → 0% (eliminated from the top 25)
RelayCacheSet.Marshal: 30 GB → 19 GB (-37%)

Tradeoff: upstream bandwidth goes up because nodes now send plain JSON on the wire for smart-router traffic. For co-located upstreams this is free; geographically-distant deployments trade CPU for bytes. Per-request deadlines are unchanged (caller context via NewRequestWithContext); the client's 5-minute DefaultHTTPTimeout is a safety backstop above any realistic upstream timeout.

Covered by:

• TestHTTPDirectRPCConnection_UsesSharedOptimizedTransport: regression lock-in that two HTTPDirectRPCConnection instances share the singleton common.SharedHttpTransport() for pooling and TLS session reuse.
• TestHTTPDirectRPCConnection_AdvertisesAcceptEncodingIdentity: httptest end-to-end assertion that both SendRequest and DoHTTPRequest send Accept-Encoding: identity on outbound requests and that response bodies are returned untransformed.

Description

Closes: #XXXX

---

Author Checklist

All items are required. Please add a note to the item if the item is not applicable and
please add links to any relevant follow up issues.

I have...

• read the contribution guide
• included the correct type prefix in the PR title, you can find examples of the prefixes below:
• confirmed ! in the type prefix if API or client breaking change
• targeted the main branch
• provided a link to the relevant issue or specification
• reviewed "Files changed" and left comments if necessary
• included the necessary unit and integration tests
• updated the relevant documentation or specification, including comments for documenting Go code
• confirmed all CI checks have passed

Reviewers Checklist

All items are required. Please add a note if the item is not applicable and please add
your handle next to the items reviewed if you only reviewed selected items.

I have...

• confirmed the correct type prefix in the PR title
• confirmed all author checklist items have been addressed
• reviewed state machine logic, API design and naming, documentation is accurate, tests and test coverage

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

Flag	Coverage Δ
consensus	8.98% <ø> (ø)
protocol	35.25% <100.00%> (+0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
protocol/lavasession/direct_rpc_connection.go	32.29% <100.00%> (+4.65%)	⬆️

... and 4 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[github-actions]

Test Results

0 tests  ±0   0 ✅ ±0   0s ⏱️ ±0s
0 suites ±0   0 💤 ±0 
7 files   ±0   0 ❌ ±0 

Results for commit 71eeb95. ± Comparison against base commit 2d92617.

♻️ This comment has been updated with latest results.