Re: questions about the async reads

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hello!

I tried to bring async reads a few years ago. The change improved
synchronous reads at the price of some penalty for randoms reads
and, of course, significant complexity grow.

To make a long story short: random read performance tends to be
far more important than the sequential one.

For more details take a look on:
  * https://github.com/ceph/ceph/pull/19380
  * https://github.com/ceph/ceph/pull/29527#issuecomment-522505628

Regards,
Radek

On Wed, Mar 23, 2022 at 9:05 AM C0reFast <topchenfu@xxxxxx> wrote:
>
> Hello everyone
>
> I've reading the ceph codes for a while, and i found that in the ReplicatedBackend, the objects_read_async function is not implenmented.
>
> ```
> void ReplicatedBackend::objects_read_async(
>   const hobject_t &hoid,
>   const list<pair<boost::tuple<uint64_t, uint64_t, uint32_t>,
>   pair<bufferlist*, Context*> > > &to_read,
>   Context *on_complete,
>   bool fast_read)
> {
>   ceph_abort_msg("async read is not used by replica pool");
> }
> ```
> in my opinion,make read ops asynchronously may get a lot performance enhance,
> so is there anyone could tell me why ceph ReplicatedBackend does't implenment the async reads?
>
> thanks!
>
> _______________________________________________
> Dev mailing list -- dev@xxxxxxx
> To unsubscribe send an email to dev-leave@xxxxxxx

_______________________________________________
Dev mailing list -- dev@xxxxxxx
To unsubscribe send an email to dev-leave@xxxxxxx




[Index of Archives]     [CEPH Users]     [Ceph Devel]     [Ceph Large]     [Information on CEPH]     [Linux BTRFS]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux