Hi Ilya
Ah, thanks. I misunderstood that part. However, I can't get it to work,
data still goes to the wrong pool.
I did this, which seemed to stick.
# ceph config set global rbd_default_data_pool rbd_data
# ceph config dump | grep rbd_default
global advanced rbd_default_data_pool
rbd_data
I turned on mirroring and image sync started alright:
# rbd mirror image status rbd/depot64
depot64:
global_id: 9315eb0f-207c-4f7c-928c-ec4ba7ba7c47
state: up+syncing
description: bootstrapping, IMAGE_SYNC/COPY_IMAGE 29%
service: dcn-ceph-02.qunlre on dcn-ceph-02
last_update: 2021-08-18 19:21:22
# rbd info depot64
rbd image 'depot64':
size 1 TiB in 262144 objects
order 22 (4 MiB objects)
snapshot_count: 2
id: bfee43c1858765
block_name_prefix: rbd_data.bfee43c1858765
format: 2
features: layering, exclusive-lock, journaling
op_features:
flags:
create_timestamp: Wed Aug 18 19:02:45 2021
access_timestamp: Wed Aug 18 19:02:45 2021
modify_timestamp: Wed Aug 18 19:02:45 2021
journal: bfee43c1858765
mirroring state: enabled
mirroring mode: journal
mirroring global id: 9315eb0f-207c-4f7c-928c-ec4ba7ba7c47
mirroring primary: false
No data in the rbd_data pool though, still going to the rbd pool:
# ceph df
--- RAW STORAGE ---
CLASS SIZE AVAIL USED RAW USED %RAW USED
hdd 224 TiB 162 TiB 61 TiB 61 TiB 27.32
ssd 1.3 TiB 860 GiB 481 GiB 481 GiB 35.89
TOTAL 225 TiB 163 TiB 62 TiB 62 TiB 27.37
--- POOLS ---
POOL ID PGS STORED OBJECTS USED %USED MAX AVAIL
device_health_metrics 1 1 47 MiB 61 140 MiB 0.02 240 GiB
cephfs_data 3 128 33 TiB 56.26M 57 TiB 36.29 50 TiB
cephfs_metadata 4 32 26 GiB 1.32M 53 GiB 6.80 360 GiB
rbd_data 5 32 8 KiB 1 16 KiB 0 50 TiB
rbd 6 128 257 GiB 62.55k 420 GiB 36.84 360 GiB
Did I miss something obvious?
Thanks,
Torkil
On 18/08/2021 14.30, Ilya Dryomov wrote:
On Wed, Aug 18, 2021 at 12:40 PM Torkil Svensgaard <torkil@xxxxxxxx> wrote:
Hi
I am looking at one way mirroring from cluster A to B cluster B.
As pr [1] I have configured two pools for RBD on cluster B:
1) Pool rbd_data using default EC 2+2
2) Pool rbd using replica 2
I have a peer relationship set up so when I enable mirroring on an image
in cluster A it will be replicated to cluster B but it will put both
data and metadata in the rbd pool.
How do I get the rbd-mirror daemon to use rbd_data for data and rbd for
metadata only?
Thanks,
Torkil
[1]
https://docs.ceph.com/en/latest/rados/operations/erasure-code/#erasure-coding-with-overwrites
Hi Torkil,
This is covered here:
https://docs.ceph.com/en/latest/rbd/rbd-mirroring/#data-pools
Thanks,
Ilya
_______________________________________________
ceph-users mailing list -- ceph-users@xxxxxxx
To unsubscribe send an email to ceph-users-leave@xxxxxxx
_______________________________________________
ceph-users mailing list -- ceph-users@xxxxxxx
To unsubscribe send an email to ceph-users-leave@xxxxxxx
