v3.0.0-alpha5 Known issue

[journeytosilius]

Known problems
There is still some instability in the RPC

What does some instability in the RPC means exactly ?

[AskAlexSharov]

please use main git branch - we fixed rpc there.
list of fixes: https://github.com/erigontech/erigon/milestone/5?closed=1

[journeytosilius]

but which release to use for erigon 3 ? currently I'm using

* (HEAD detached at v3.0.0-alpha5)
  main

And it synced ok, but ws RPC is very slow ? My backend makes concurrent RPC calls and before, it was fast ... now it looks like it uses all threads on the server and it's very slow ? Maybe need to tune the throughput, but how do I know the variables to make it work like default geth ? - I will open a new issue

[AskAlexSharov]

@journeytosilius

• use main
• before - when?
• Erigon3 doesn't store receipts/logs for now (Erigon2 did). Because re-exec of 1 historical txn is cheaper in E3 than in E2. We still adding some optimizations to RPC - to make re-exec of 1 txn cheaper. If after our work done - perf of fetching receipts will still be not enough: we will introduce "optional persistence" of rpc logs.
• some flags: --rpc.batch.concurrency --rpc.batch.limit --db.read.concurrency

[journeytosilius]

@AskAlexSharov this was just me first testing v3.0.0-alpha5
Then I realized that this was the problem and I made another issue : #12892 (comment),
I forgot that Erigon 3 is still alpha, nevermind :)