Hi,
Has there been any progress on this issue ? is there quick recover method? I
have same problem with you that first 4k block of osd metadata is invalid. It
will pay a heavy price to recreate osd.
Thanks.
___
ceph-users mailing list --
Hello,
There is no such usage collected today, see [1] and [2] – where [2] is a
specification on how
one community member wanted to implement the feature but nobody has put in the
work yet
that we know of.
[1] https://tracker.ceph.com/issues/47342
[2] https://tracker.ceph.com/issues/54972
Let's take for example a situation where I have a standart storage class backed by HDDs and a fast one on SSDs. The user will mix the classes in the bucket and I would like to know how much space Is he taking on the HDDs and how much on the SSDs so I can bill him.In this scenario I don't care that
A bucket may contain objects spread across multiple storage classes, and AIUI
the head object is always in the default storage class, so I'm not sure
*exactly* what you're after here.
> On Apr 4, 2024, at 17:09, Ondřej Kukla wrote:
>
> Hello,
>
> I’m playing around with Storage classes in
Hello,
I’m playing around with Storage classes in rgw and I’m looking for ways to see
per bucket statistics for the diferent storage classes (for billing purposes
etc.).
I though that I would add another object to the bucket usage response like for
multiparts - rgw.multimeta, but it’s
Sorry to keep asking for more info, but can I also get what `cephadm
gather-facts` on that host returns for "memory_total_kb". Might end up
creating a unit test out of this case if we have a calculation bug here.
On Thu, Apr 4, 2024 at 4:05 PM Mads Aasted wrote:
> sorry for the double send,
sorry for the double send, forgot to hit reply all so it would appear on
the page
Hi Adam
If we multiply by 0.7, and work through the previous example from that
number, we would still arrive at roughly 2.5 gb for each osd. And the host
in question is trying to set it to less than 500mb.
I have
Thank you again Casey for putting us on the right track regarding the
changes in multisite resharding support.
When going through the various changelogs we didn't pay too much
attention to those changes as this cluster doesn't use any multisite
features.
We have now upgraded to Reef and all
Initial indication shows "osd_async_recovery_min_cost = 0" to be a huge
win here. Some initial thoughts. Were this not for the fact that the index
(and other OMAP pools) were isolated to their own OSDs in this cluster this
tunable would seemingly cause data/blob objects from data pools to async
On Thu, Apr 4, 2024 at 8:23 AM Anthony D'Atri wrote:
>
> Network RTT?
No, it's sadly not that clever. There's a crush_location configurable
that you can set on clients (to a host, or a datacenter, or any other
CRUSH bucket), and as long as part of it matches the CRUSH map then it
will feed IOs
Not a requirement but it makes it a LOT faster.
> On Apr 4, 2024, at 03:54, Szabo, Istvan (Agoda)
> wrote:
>
> Hi,
>
> Let's say thin provisioned and no, no fast-diff and object map enabled. As I
> see this is a requirements to be able to use "du".
>
>
> Istvan Szabo
> Staff Infrastructure
Network RTT?
> On Apr 4, 2024, at 03:44, Noah Elias Feldt wrote:
>
> Hello,
> I have a question about a setting for RBD.
> How exactly does "rbd_read_from_replica_policy" with the value "localize"
> work?
> According to the RBD documentation, read operations will be sent to the
> closest OSD
First, I guess I would make sure that peon7 and peon12 actually could pass
the host check (you can run "cephadm check-host" on the host directly if
you have a copy of the cephadm binary there) Then I'd try a mgr failover
(ceph mgr fail) to clear out any in memory host values cephadm might have
and
Hi Boris,
thank you for your answer.
What i really did not expect is, that purging an OSD leads to yet another
rebalancing, after already having it destroyed(!), after it was taken out and
after the rebalancing seemed to have been completed.
So with the next host - i have some more to remove
Hi Tobias,
what we usually do, when we want to remove an OSD is to reweight the
crush map to 0. This stops the rebalancing after removing the OSD from
the crush map. Setting an OSD to out, keeps it weighted in the crush
map and when it gets removed, the cluster will rebalance the PGs to
reflect
Dear Cephers,
reorganizing one of our clusters i'm removing some hosts from it, taking "out"
all OSDs on these hosts and waiting until all PGs are fine.
After stopping and destroying all OSDs on one host i notice, that "purge" of
such destroyed OSDs temporarily leads to degraded and misplaced
Hi,
I’ve added some new nodes to our Ceph cluster. Only did the host add, had not
added the OSD’s yet.
Due to a configuration error I had to reinstall some of them. But I forgot to
remove the nodes from Ceph first. I did a “ceph orch host rm peon7 --offline
—force” before re-adding them to the
Thank you, Eugen. This makes sense.
/Z
On Thu, 4 Apr 2024 at 10:32, Eugen Block wrote:
> Hi,
>
> the noin flag seems to be only applicable to existing OSDs which are
> already in the crushmap. It doesn't apply to newly created OSDs, I
> could confirm that in a small test cluster with Pacific
Thanks, this is a good suggestion!
/Z
On Thu, 4 Apr 2024 at 10:29, Janne Johansson wrote:
> Den tors 4 apr. 2024 kl 06:11 skrev Zakhar Kirpichenko :
> > Any comments regarding `osd noin`, please?
> > >
> > > I'm adding a few OSDs to an existing cluster, the cluster is running
> with
> > > `osd
Hi,
Let's say thin provisioned and no, no fast-diff and object map enabled. As I
see this is a requirements to be able to use "du".
Istvan Szabo
Staff Infrastructure Engineer
---
Agoda Services Co., Ltd.
e:
Hello,
I have a question about a setting for RBD.
How exactly does "rbd_read_from_replica_policy" with the value "localize" work?
According to the RBD documentation, read operations will be sent to the closest
OSD as determined by the CRUSH map. How does the client know exactly which OSD
I am
Hi,
the noin flag seems to be only applicable to existing OSDs which are
already in the crushmap. It doesn't apply to newly created OSDs, I
could confirm that in a small test cluster with Pacific and Reef. I
don't have any insights if that is by design or not, I assume it's
supposed to
Den tors 4 apr. 2024 kl 06:11 skrev Zakhar Kirpichenko :
> Any comments regarding `osd noin`, please?
> >
> > I'm adding a few OSDs to an existing cluster, the cluster is running with
> > `osd noout,noin`:
> >
> > cluster:
> > id: 3f50555a-ae2a-11eb-a2fc-ffde44714d86
> > health:
23 matches
Mail list logo