> 26 марта 2021 г., в 11:00, Andrey Borodin <x4...@yandex-team.ru> написал(а):
>
>> I'm not saying the 0002 patch is bug-free yet though, it's a bit finickity.
> I think the idea of speeding up linear search is really really good for
> scaling SLRUs. It's not even about improving normal performance of the
> cluster, but it's important from preventing pathological degradation under
> certain circumstances. Bigger cache really saves SLAs :) I'll look into the
> patch more closely this weekend. Thank you!
Some thoughts on HashTable patch:
1. Can we allocate bigger hashtable to reduce probability of collisions?
2. Can we use specialised hashtable for this case? I'm afraid hash_search()
does comparable number of CPU cycles as simple cycle from 0 to 128. We could
inline everything and avoid hashp->hash(keyPtr, hashp->keysize) call. I'm not
insisting on special hash though, just an idea.
3. pageno in SlruMappingTableEntry seems to be unused.
Thanks!
Best regards, Andrey Borodin.