Hi,
if the OSDs are not too full it's probably the crush weight of those
hosts and OSDs. CRUSH tries to distribute the data evenly to all three
hosts because they have the same weight (9.31400). But since two OSDs
are missing the distribution doesn't finish. If you can't replace the
Hi Philipp,
Given 256 PG's triple replicated onto 4 OSD's you might be encountering
the "PG overdose protection" of OSDs. Take a look at 'ceph osd df' and
see the number of PG's that are mapped to each OSD (last column or near
the last). The default limit is 200, so if any OSD exceeds that,
hi,
On 2019-10-22 08:05, Eugen Block wrote:
> Hi,
>
> can you share `ceph osd tree`? What crush rules are in use in your
> cluster? I assume that the two failed OSDs prevent the remapping because
> the rules can't be applied.
>
ceph osd tree gives:
ID WEIGHT TYPE NAMEUP/DOWN
Hi,
can you share `ceph osd tree`? What crush rules are in use in your
cluster? I assume that the two failed OSDs prevent the remapping
because the rules can't be applied.
Regards,
Eugen
Zitat von Philipp Schwaha :
hi,
I have a problem with a cluster being stuck in recovery after osd
hi,
I have a problem with a cluster being stuck in recovery after osd
failure. at first recovery was doing quite well, but now it just sits
there without any progress. I currently looks like this:
health HEALTH_ERR
36 pgs are stuck inactive for more than 300 seconds