Re: Slow delete speed through the s3 API

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

 



Is your client using the DeleteObjects call to delete 1000 per request?:
https://docs.aws.amazon.com/AmazonS3/latest/API/API_DeleteObjects.html

On Fri, Jun 3, 2022 at 9:35 AM J-P Methot <jp.methot@xxxxxxxxxxxxxxxxx>
wrote:

> Read/writes are super fast. It's only deletes that are incredibly slow,
> both through the s3 api and radosgw commands. It doesn't have a lot of
> index shards with 11 and it basically has one bucket with huge objects.
> We're trying to store backups in there, essentially. New backups write
> themselves about 100 times faster than old backups can be deleted.
>
> On 6/2/22 19:15, Wesley Dillingham wrote:
> > Is it just your deletes which are slow or writes and read as well?
> >
> > On Thu, Jun 2, 2022, 4:09 PM J-P Methot <jp.methot@xxxxxxxxxxxxxxxxx>
> > wrote:
> >
> >     I'm following up on this as we upgraded to Pacific 16.2.9 and deletes
> >     are still incredibly slow. The pool rgw is using is a fairly small
> >     erasure coding pool set at 8 + 3. Is there anyone who's having the
> >     same
> >     issue?
> >
> >     On 5/16/22 15:23, J-P Methot wrote:
> >     > Hi,
> >     >
> >     > First of all, a quick google search shows me that questions
> >     about the
> >     > s3 API slow object deletion speed have been asked before and are
> >     well
> >     > documented. My issue is slightly different, because I am getting
> >     > abysmal speeds of 11 objects/second on a full SSD ceph running
> >     Octopus
> >     > with about a hundred OSDs. This is much lower than the Redhat
> >     reported
> >     > limit of 1000 objects/second.
> >     >
> >     > I've seen elsewhere that it was a Rocksdb limitation and that it
> >     would
> >     > be fixed in Pacific, but the Pacific release logs do not show me
> >     > anything that suggest that. Furthermore, I have limited control
> >     over
> >     > the s3client deleting the files as it's a 3rd-party open source
> >     > automatic backup program.
> >     >
> >     > Could updating to Pacific fix this issue? Is there any
> >     configuration
> >     > change I could do to speed up object deletion?
> >     >
> >     --
> >     Jean-Philippe Méthot
> >     Senior Openstack system administrator
> >     Administrateur système Openstack sénior
> >     PlanetHoster inc.
> >
> >     _______________________________________________
> >     ceph-users mailing list -- ceph-users@xxxxxxx
> >     To unsubscribe send an email to ceph-users-leave@xxxxxxx
> >
> --
> Jean-Philippe Méthot
> Senior Openstack system administrator
> Administrateur système Openstack sénior
> PlanetHoster inc.
> _______________________________________________
> ceph-users mailing list -- ceph-users@xxxxxxx
> To unsubscribe send an email to ceph-users-leave@xxxxxxx
>
_______________________________________________
ceph-users mailing list -- ceph-users@xxxxxxx
To unsubscribe send an email to ceph-users-leave@xxxxxxx




[Index of Archives]     [Information on CEPH]     [Linux Filesystem Development]     [Ceph Development]     [Ceph Large]     [Ceph Dev]     [Linux USB Development]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [xfs]


  Powered by Linux