Le 20/09/2023 à 08:50, Stefan Kooman a écrit :
On 20-09-2023 08:44, Patrick Begou wrote:
Hi Patrick,
I agree that learning Ceph today with Octopus is not a good idea,
but, as a newbie with this tool, I was not able to solve the HDD
detection problem and my post about it on this forum do not provide
any help
(https://lists.ceph.io/hyperkitty/list/ceph-users@xxxxxxx/message/OPMWHJ4ZFCOOPUY6ST4WAJ4G4ASJFALM/).
I've also looked for a list of new unsupported hardware between
Octopus and Pacific without success. I've also received a private
mail from a Sweden user reading the forum last week and having the
same HDD detection problem with 17.2.6. He was asking if I have
solved it. He tells me he will try to debug.
In my mind, an old version of Ceph on an old material had more
chances to be stable and bug free too.
Yes I have created file systems (datacfs) and I can create a snapshot
by hand using cephadm. I've just tested:
# ceph fs set datacfs allow_new_snaps true
# ceph-fuse /mnt
# mkdir /mnt/.snap/$(TZ=CET date +%Y-%m-%d:%H-%M-%S)
and I have a snapshot. I can remove it too.
May be today my goal should be:
1- try to undo "ceph mgr module enable snap_schedule --force"
(always a bad idea in my mind to use options like "--force")
2- launch update to nautilus now that all HDDs are configured. In my
Ceph learning process there is also the step to test update procedures.
That would be Pacific, as Nautilus is the release before Octopus.
Good luck,
Gr. Stefan
Yes, of course Stefan. I need an extra cup of coffee this morning 😁
Patrick
_______________________________________________
ceph-users mailing list -- ceph-users@xxxxxxx
To unsubscribe send an email to ceph-users-leave@xxxxxxx
