Hide Forgot
Additional info: Document replicating or mirroring a Ceph Block Device image on one cluster/pool to another cluster/pool, primarily for disaster recovery.
Following are my observations while verifying the doc, Please modify the doc as mentioned below.. In the Configure Mirroring Section : https://access.qa.redhat.com/documentation/en/red-hat-ceph-storage/2/block-device-guide/#configuring_mirroring 1. => One block device client is connected to each of the clusters: client.master and client.slave. For information on installing Ceph clients, see the Client Installation chapter in the Installation Guide for Red Hat Enterprise Linux. -- the keyring should be in the format "master.client.master.keyring" - with only client.master and client.slave rbd fails with cephx authentication. We need to create a new cephx authentication. 2. => On both client hosts, install the rbd-mirror package: # yum install rbd-mirror -- We have to install "ceph-common" and also copy the admin keyring to the Client host. Please mention about that. Update the same for Image configuration also 3. => Add the clusters as peers: $ rbd mirror pool peer add data client.slave@slave --cluster master $ rbd mirror pool peer add data client.master@master --cluster slave -- This section should be moved after enabling the pool mirroring This fails as the pool needs to be enabled first. Steps 4 & 5 should to be interchanged both in Pool and Image configurations. 4. => Configuring Image Mirroring On both client hosts, enable and start the rbd-mirror daemon: -- Add one more command to make the daemon start by itself after rebooting # systemctl enable ceph-rbd-mirror.target = > The stack pool on the master cluster contains the image1 and image2. Journaling is enabled on the images. See Enabling Journaling for details. -- The name of the pool mentioned in prereq is "stack" but the syntax and examples contains pool name as "data".
Looks good to me.. Moving to Verified state.