add microbatching to proxy forward loop #206
Conversation
add 40ms between next message in channel and draining the channel
| let mut transactions_batch: Vec<Vec<u8>> = packet.transactions.clone(); | ||
|
|
||
| 'more: while let Ok(more) = per_connection_receiver.try_recv() { | ||
| 'drain_loop: while let Ok(more) = per_connection_receiver.try_recv() { |
There was a problem hiding this comment.
why do you need to specify drain_loop, seems like goto which is not required right ?
There was a problem hiding this comment.
it's not needed but reads more clearly IMO
There was a problem hiding this comment.
imo this is a bad design. if u need labels to break out loops. could be done better
There was a problem hiding this comment.
no I do not need them - it just reads better!
compare it with this - when I scan that code I constantly look for the scope of breaks:
break+continue are so hard to read because they are so contextual IMO - this can be solved by labeling the scope
There was a problem hiding this comment.
I was staring a bit on the code, but so little room for improvement - can you suggest something?
There was a problem hiding this comment.
I guess I read more clearly when i read continue and break logic.
|
Max advices to not do that because transactions are MTU sized |
add 40ms between next message in channel and draining the channel; limit the batch size to 50
this chart (metric literpcproxy_batch_size) shows the difference between no microbatching and 40ms: batch size of 1 is without microbatching: