itrofimow

Hm, i'm pretty sure that it was still `NonFifoSpscQueue` when i created the issue, however your commit seems to be pushed 10 hours earlier – is that an artifact of...

I've got a draft that works (unless it crashes randomly upon hard to track down races 😉), which i expect to polish enough to PR in a week or so....

So i took some time to proof the point: Applied this [patch](https://pastebin.com/raw/URQvAmCx) to samples/hello_service and ran some http requests. 1. This is without background task, expected: ``` itrofimow@itrofimow-RedmiBook-16:~$ ~/Downloads/bombardier-linux-amd64 -c...

> It'll help us a lot in a few months, when we start the scheduling customization feature That's nice to hear, glad to help) However i think the problem with...

> Just a few nitpicks for now. Will take a deeper look later This PR is still a work in progress (it's even draft here on github) and it has...

> We've updated the moodycamel. > > Now we have three different opinions on MAX_SEMA_SPINS proper value: 0|1, 16, default|10000. We need to do some experiments on production service... I...

> Looks like it is not fixed yet, the fix is still not merged [boostorg/lockfree@8c76094](https://github.com/boostorg/lockfree/commit/8c760949d7edc21490f97f581dc821d0e9711308) The commit you mentioned is not the one i'm referencing, and although it's not bound...

> or the current behavior does harm in some practical case? Well, depends on what one would consider 'harmful', but i would argue that the current behavior it not consistent...

@apolukhin bump

Seems like it's a WIP already https://github.com/ClickHouse/clickhouse-cpp/pull/219