'ceph osd crush tunables optimal'
or adjust an offline map file via the crushtool command line (more
annoying) and retest; I suspect that is the problem.
http://ceph.com/docs/master/rados/operations/crush-map/#tunables
That solves the bug with weight 0, thanks.
But is still get the
On Mon, 6 Jan 2014, Dietmar Maurer wrote:
'ceph osd crush tunables optimal'
or adjust an offline map file via the crushtool command line (more
annoying) and retest; I suspect that is the problem.
http://ceph.com/docs/master/rados/operations/crush-map/#tunables
That solves the bug
Host with only one osd gets too much data.
I think this is just fundamentally a problem with distributing 3 replicas
over only 4
hosts. Every piece of data in the system needs to include either host 3 or 4
(and
thus device 4 or 5) in order to have 3 replicas (on separate hosts). Add
I think this is just fundamentally a problem with distributing 3
replicas over only 4 hosts. Every piece of data in the system needs
to include either host 3 or 4 (and thus device 4 or 5) in order to
have 3 replicas (on separate hosts). Add more hosts or disks and the
distribution will
Run
'ceph osd crush tunables optimal'
or adjust an offline map file via the crushtool command line (more
annoying) and retest; I suspect that is the problem.
http://ceph.com/docs/master/rados/operations/crush-map/#tunables
sage
On Fri, 3 Jan 2014, Dietmar Maurer wrote:
In both cases, you
I try to understand the default crush rule:
rule data {
ruleset 0
type replicated
min_size 1
max_size 10
step take default
step chooseleaf firstn 0 type host
step emit
}
Is this the same as:
rule data {
ruleset 0
type
On 01/02/2014 10:40 AM, Dietmar Maurer wrote:
I try to understand the default crush rule:
rule data {
ruleset 0
type replicated
min_size 1
max_size 10
step take default
step chooseleaf firstn 0 type host
step emit
}
Is this the same
On Thu, 2 Jan 2014, Wido den Hollander wrote:
On 01/02/2014 10:40 AM, Dietmar Maurer wrote:
I try to understand the default crush rule:
rule data {
ruleset 0
type replicated
min_size 1
max_size 10
step take default
On Thu, 2 Jan 2014, Dietmar Maurer wrote:
iirc, chooseleaf goes down the tree and descents into multiple leafs
to find what you are looking for.
choose goes into that leaf and tries to find what you are looking for
without going into subtrees.
Right. To a first approximation,
iirc, chooseleaf goes down the tree and descents into multiple leafs
to find what you are looking for.
choose goes into that leaf and tries to find what you are looking for
without going into subtrees.
Right. To a first approximation, these rules are equivalent. The difference
is
The other difference is if you have one of the two OSDs on the host marked
out.
In the choose case, the remaining OSD will get allocated 2x the data; in the
chooseleaf case, usage will remain proportional with the rest of the cluster
and
the data from the out OSD will be distributed across
--test IIRC. Yep!
sage
On Thu, 2 Jan 2014, Dietmar Maurer wrote:
The other difference is if you have one of the two OSDs on the host marked
out.
In the choose case, the remaining OSD will get allocated 2x the data; in the
chooseleaf case, usage will remain proportional with the rest
In both cases, you only get 2 replicas on the remaining 2 hosts.
OK, I was able to reproduce this with crushtool.
The difference is if you have 4 hosts with 2 osds. In the choose case, you
have
some fraction of the data that chose the down host in the first step (most of
the
attempts,
I also don't really understand why crush selects OSDs with weight=0
host prox-ceph-3 {
id -4 # do not change unnecessarily
# weight 3.630
alg straw
hash 0 # rjenkins1
item osd.4 weight 0
}
root default {
id -1 # do not
14 matches
Mail list logo
