search

beethovenxfi / beethovenx-backend

MIT License
13 stars 15 forks source link

analyze SOR results #625

Open franzns opened 10 months ago

franzns commented 10 months ago

Need to analyze the collected SOR metrics. Is the new SOR faster? Is the new SOR performant in terms of CPU/MEM? Do we need to decouple? How do they compare for swap performance? (it's expected that v1 is better, but how much worse is v2?)

Strange routes: 1 wFTM>BEETS (3.9% price impact, through 6 pools) 1 COMB>BEETS (convoluted route, could be simpler with better price)

check whether this is sor v1 or v2

franzns commented 10 months ago

above results are from sor v2, sor v1 broke with DB schema changes we did before christmast...

need to investigate why it routes through such strange pools for 1wftm->beets