EC profile datastore usage - question
Hello Community, I would like to ask about help in explanation situation. There is Rados gateway with EC pool profile k=6 m=4. So it shoud take something about 1.4 - 2.0 data usage more from raw data if I’m correct. Rados df shows me: 116 TiB used and WR 26 TiB Can You explain this? It is about 4.5*WR used data. Why? Regards Mateusz Skała
Can you post the output of a couple of commands: ceph df ceph osd pool ls detail Then we can probably explain the utilization you're seeing. -----Original Message----- From: Mateusz Skała <mateusz.skala@gmail.com> Sent: Saturday, July 18, 2020 1:35 AM To: ceph-users@ceph.io Subject: [ceph-users] EC profile datastore usage - question Hello Community, I would like to ask about help in explanation situation. There is Rados gateway with EC pool profile k=6 m=4. So it shoud take something about 1.4 - 2.0 data usage more from raw data if I’m correct. Rados df shows me: 116 TiB used and WR 26 TiB Can You explain this? It is about 4.5*WR used data. Why? Regards Mateusz Skała _______________________________________________ ceph-users mailing list -- ceph-users@ceph.io To unsubscribe send an email to ceph-users-leave@ceph.io
Hi Mateusz, I think you might be hit by: https://tracker.ceph.com/issues/44213 This is fixed in upcoming Pacific release. Nautilus/Octopus backport is under discussion for now. Thanks, Igor On 7/18/2020 8:35 AM, Mateusz Skała wrote:
Hello Community, I would like to ask about help in explanation situation. There is Rados gateway with EC pool profile k=6 m=4. So it shoud take something about 1.4 - 2.0 data usage more from raw data if I’m correct. Rados df shows me: 116 TiB used and WR 26 TiB Can You explain this? It is about 4.5*WR used data. Why? Regards Mateusz Skała _______________________________________________ ceph-users mailing list -- ceph-users@ceph.io To unsubscribe send an email to ceph-users-leave@ceph.io
Hi Igor, Given the patch histories and the rejection of the previous patch for the one in favor of defaulting to 4k block size, does this essentially mean ceph does not support higher block sizes when using erasure coding? Will the ceph project be updating their documentation and references to let everyone know that larger blocks don't interact with EC pools as intended? Sincerely, On Mon, Jul 20, 2020 at 9:06 AM Igor Fedotov <ifedotov@suse.de> wrote:
Hi Mateusz,
I think you might be hit by:
https://tracker.ceph.com/issues/44213
This is fixed in upcoming Pacific release. Nautilus/Octopus backport is under discussion for now.
Thanks,
Igor
On 7/18/2020 8:35 AM, Mateusz Skała wrote:
Hello Community, I would like to ask about help in explanation situation. There is Rados gateway with EC pool profile k=6 m=4. So it shoud take something about 1.4 - 2.0 data usage more from raw data if I’m correct. Rados df shows me: 116 TiB used and WR 26 TiB Can You explain this? It is about 4.5*WR used data. Why? Regards Mateusz Skała _______________________________________________ ceph-users mailing list -- ceph-users@ceph.io To unsubscribe send an email to ceph-users-leave@ceph.io
ceph-users mailing list -- ceph-users@ceph.io To unsubscribe send an email to ceph-users-leave@ceph.io
-- Steven Pine webair.com *P* 516.938.4100 x *E * steven.pine@webair.com <https://www.facebook.com/WebairInc/> <https://www.linkedin.com/company/webair>
Hi Steven, IMO your statement about "not supporting higher block sizes" is too strong. From my experience excessive space usage for EC pools tend to depend on write access pattern, input block sizes and/or object sizes. Hence I'm pretty sure this issue isn't present/visible for every cluster with alloc size > 4K and EC pool enabled. E.g. one wouldn't notice it when the majority of objects are large enough and/or written just once. But honestly I haven't performed that comprehensive investigation on the topic. And perhaps you're right and we should cover the issue in the documentation. Please feel free to fire corresponding ticket in Ceph tracker... Thanks, Igor On 7/20/2020 7:02 PM, Steven Pine wrote:
Hi Igor,
Given the patch histories and the rejection of the previous patch for the one in favor of defaulting to 4k block size, does this essentially mean ceph does not support higher block sizes when using erasure coding? Will the ceph project be updating their documentation and references to let everyone know that larger blocks don't interact with EC pools as intended?
Sincerely,
On Mon, Jul 20, 2020 at 9:06 AM Igor Fedotov <ifedotov@suse.de <mailto:ifedotov@suse.de>> wrote:
Hi Mateusz,
I think you might be hit by:
https://tracker.ceph.com/issues/44213
This is fixed in upcoming Pacific release. Nautilus/Octopus backport is under discussion for now.
Thanks,
Igor
On 7/18/2020 8:35 AM, Mateusz Skała wrote: > Hello Community, > I would like to ask about help in explanation situation. > There is Rados gateway with EC pool profile k=6 m=4. So it shoud take > something about 1.4 - 2.0 data usage more from raw data if I’m correct. > Rados df shows me: > 116 TiB used and WR 26 TiB > Can You explain this? It is about 4.5*WR used data. Why? > Regards > Mateusz Skała > _______________________________________________ > ceph-users mailing list -- ceph-users@ceph.io <mailto:ceph-users@ceph.io> > To unsubscribe send an email to ceph-users-leave@ceph.io <mailto:ceph-users-leave@ceph.io> _______________________________________________ ceph-users mailing list -- ceph-users@ceph.io <mailto:ceph-users@ceph.io> To unsubscribe send an email to ceph-users-leave@ceph.io <mailto:ceph-users-leave@ceph.io>
-- Steven Pine webair.com <http://webair.com>
*P* 516.938.4100 x
*E * steven.pine@webair.com <mailto:steven.pine@webair.com>
<https://www.facebook.com/WebairInc/> <https://www.linkedin.com/company/webair>
participants (4)
-
Eric Smith
-
Igor Fedotov
-
Mateusz Skała
-
Steven Pine