Compress editoast cached data

[younesschrifi]

The json request and response from the pathfinding and the simulation take too much time to be get from valkey or push in valkey. The solution is to use gzip to compress the payloads.

Benchmarks

All benchmark were run on dev.

No compression (BATCH_SIZE=500)

• LMR: ~20s
• Get from cache (total): 9.4s
• Get from cache (sum): 14.0s

Zstd -1 (BATCH_SIZE=500)

• LMR: ~20s
• Get from cache (total): 13.7s
• Get from cache (sum): 22.6s

LZ4 -1 (BATCH_SIZE=500)

• LMR: ~14.4s
• Get from cache (total): 8.8s
• Get from cache (sum): 11.9s

[codecov-commenter]

⚠️ Please install the to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

Attention: Patch coverage is 79.48718% with 8 lines in your changes missing coverage. Please review.

Project coverage is 81.81%. Comparing base (4f2c918) to head (b0edd25).
Report is 37 commits behind head on dev.

Files with missing lines	Patch %	Lines
editoast/src/valkey_utils.rs	77.77%	8 Missing ⚠️

❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@           Coverage Diff            @@
##              dev   #10527    +/-   ##
========================================
  Coverage   81.81%   81.81%            
========================================
  Files        1073     1073            
  Lines      106718   106981   +263     
  Branches      730      728     -2     
========================================
+ Hits        87307    87528   +221     
- Misses      19372    19414    +42     
  Partials       39       39            

Flag	Coverage Δ
editoast	74.16% <79.48%> (-0.14%)	⬇️
front	89.37% <ø> (+0.02%)	⬆️
gateway	2.18% <ø> (ø)
osrdyne	3.28% <ø> (ø)
railjson_generator	87.50% <ø> (ø)
tests	88.14% <ø> (+1.09%)	⬆️

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

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[woshilapin]

Hopes this improve what we're looking for. Do you have some numbers? It would be nice to put them in the description so future us can find these when reevaluating the situation.

[flomonster]

Hopes this improve what we're looking for. Do you have some numbers? It would be nice to put them in the description so future us can find these when reevaluating the situation.

We're benchmarking PR. I will add numbers in live :)

[leovalais]

Wouldn't reducing the batch size improve the performances since transport is likely the limiting factor here?

[flomonster]

We can merge this PR. We've made a slight improvement. The result is not very visible because of other problems I've listed here.