Hi folks:
I've been playing around with the new Ceph orchestrator and I've been running
into an interesting limitation. I have not found any documented process or
combination of things to do which can allow me to 're-adopt' cephadm deployed
OSDs.
Think use case of reinstalling operating
On Sat, Sep 26, 2020 at 04:50:34PM +, t...@postix.net wrote:
> Hi all,
>
> For those who use encryption on your OSDs, what effect do you see on
> your NVMe, SSD and HDD vs non-encrypted OSDs? I tried to find some
> info on this subject but there isn't much detail available.
>
> >From
Hi all,
For those who use encryption on your OSDs, what effect do you see on your NVMe,
SSD and HDD vs non-encrypted OSDs? I tried to find some info on this subject
but there isn't much detail available.
>From experience, dmcrypt is CPU-bound and becomes a bottleneck when used on
>very fast
Without knowing the source code and just from my observations, I would say
everytime the osd map changes, the crush/pgmap tries to fix that. However a
running backfill is not stopped and only backfill_wait would be
reconsidered.
--
Martin Verges
Managing director
Mobile: +49 174 9335695
E-Mail:
When I add an osd rebalancing is taking place, lets say ceph relocates
40 pg's.
When I add another osd during rebalancing, when ceph has only relocated
10 pgs and has to do still 30 pgs.
What happens then:
1. Is ceph just finishing the relocation of these 30 pgs and then
calculates how the
http://garmin-com-express.support/
http://webrootloginn.com/
http://www.123hpcom.de/
https://123hpcomsetup.us/
http://nortonlogins.com/
http://getgarminexpress.com/
https://canoncomijsetup.me/
___
ceph-users mailing list -- ceph-users@ceph.io
To