RBD Mirroring

RBD images can be asynchronously mirrored between two Ceph clusters. This capability is available in two modes:

  • Journal-based: This mode uses the RBD journaling image feature to ensure point-in-time, crash-consistent replication between clusters. Every write to the RBD image is first recorded to the associated journal before modifying the actual image. The remote cluster will read from this associated journal and replay the updates to its local copy of the image. Since each write to the RBD image will result in two writes to the Ceph cluster, expect write latencies to nearly double while using the RBD journaling image feature.

  • Snapshot-based: This mode uses periodically scheduled or manually created RBD image mirror-snapshots to replicate crash-consistent RBD images between clusters. The remote cluster will determine any data or metadata updates between two mirror-snapshots and copy the deltas to its local copy of the image. With the help of the RBD fast-diff image feature, updated data blocks can be quickly determined without the need to scan the full RBD image. Since this mode is not as fine-grained as journaling, the complete delta between two snapshots will need to be synced prior to use during a failover scenario. Any partially applied set of deltas will be rolled back at moment of failover.

Note

journal-based mirroring requires the Ceph Jewel release or later; snapshot-based mirroring requires the Ceph Octopus release or later.

Mirroring is configured on a per-pool basis within peer clusters and can be configured on a specific subset of images within the pool. You can also mirror all images within a given pool when using journal-based mirroring. Mirroring is configured using the rbd command. The rbd-mirror daemon is responsible for pulling image updates from the remote peer cluster and applying them to the image within the local cluster.

Depending on the desired needs for replication, RBD mirroring can be configured for either one- or two-way replication:

  • One-way Replication: When data is only mirrored from a primary cluster to a secondary cluster, the rbd-mirror daemon runs only on the secondary cluster.

  • Two-way Replication: When data is mirrored from primary images on one cluster to non-primary images on another cluster (and vice-versa), the rbd-mirror daemon runs on both clusters.

Important

Each instance of the rbd-mirror daemon must be able to connect to both the local and remote Ceph clusters simultaneously (i.e. all monitor and OSD hosts). Additionally, the network must have sufficient bandwidth between the two data centers to handle mirroring workload.

Pool Configuration

The following procedures demonstrate how to perform the basic administrative tasks to configure mirroring using the rbd command. Mirroring is configured on a per-pool basis.

These pool configuration steps should be performed on both peer clusters. These procedures assume that both clusters, named “site-a” and “site-b”, are accessible from a single host for clarity.

See the rbd manpage for additional details of how to connect to different Ceph clusters.

Note

The cluster name in the following examples corresponds to a Ceph configuration file of the same name (e.g. /etc/ceph/site-b.conf). See the ceph-conf documentation for how to configure multiple clusters. Note that rbd-mirror does not require the source and destination clusters to have unique internal names; both can and should call themselves ceph. The config files that rbd-mirror needs for local and remote clusters can be named arbitrarily, and containerizing the daemon is one strategy for maintaining them outside of /etc/ceph to avoid confusion.

Enable Mirroring

To enable mirroring on a pool with rbd, issue the mirror pool enable subcommand with the pool name, the mirroring mode, and an optional friendly site name to describe the local cluster:

rbd mirror pool enable [--site-name {local-site-name}] {pool-name} {mode}

The mirroring mode can either be image or pool:

  • image: When configured in image mode, mirroring must explicitly enabled on each image.

  • pool (default): When configured in pool mode, all images in the pool with the journaling feature enabled are mirrored.

For example:

$ rbd --cluster site-a mirror pool enable --site-name site-a image-pool image
$ rbd --cluster site-b mirror pool enable --site-name site-b image-pool image

The site name can also be specified when creating or importing a new bootstrap token.

The site name can be changed later using the same mirror pool enable subcommand but note that the local site name and the corresponding site name used by the remote cluster generally must match.

Disable Mirroring

To disable mirroring on a pool with rbd, specify the mirror pool disable command and the pool name:

rbd mirror pool disable {pool-name}

When mirroring is disabled on a pool in this way, mirroring will also be disabled on any images (within the pool) for which mirroring was enabled explicitly.

For example:

$ rbd --cluster site-a mirror pool disable image-pool
$ rbd --cluster site-b mirror pool disable image-pool

Bootstrap Peers

In order for the rbd-mirror daemon to discover its peer cluster, the peer must be registered and a user account must be created. This process can be automated with rbd and the mirror pool peer bootstrap create and mirror pool peer bootstrap import commands.

To manually create a new bootstrap token with rbd, issue the mirror pool peer bootstrap create subcommand, a pool name, and an optional friendly site name to describe the local cluster:

rbd mirror pool peer bootstrap create [--site-name {local-site-name}] {pool-name}

The output of mirror pool peer bootstrap create will be a token that should be provided to the mirror pool peer bootstrap import command. For example, on site-a:

$ rbd --cluster site-a mirror pool peer bootstrap create --site-name site-a image-pool
eyJmc2lkIjoiOWY1MjgyZGItYjg5OS00NTk2LTgwOTgtMzIwYzFmYzM5NmYzIiwiY2xpZW50X2lkIjoicmJkLW1pcnJvci1wZWVyIiwia2V5IjoiQVFBUnczOWQwdkhvQmhBQVlMM1I4RmR5dHNJQU50bkFTZ0lOTVE9PSIsIm1vbl9ob3N0IjoiW3YyOjE5Mi4xNjguMS4zOjY4MjAsdjE6MTkyLjE2OC4xLjM6NjgyMV0ifQ==

To manually import the bootstrap token created by another cluster with rbd, specify the mirror pool peer bootstrap import command, the pool name, a file path to the created token (or ‘-’ to read from standard input), along with an optional friendly site name to describe the local cluster and a mirroring direction (defaults to rx-tx for bidirectional mirroring, but can also be set to rx-only for unidirectional mirroring):

rbd mirror pool peer bootstrap import [--site-name {local-site-name}] [--direction {rx-only or rx-tx}] {pool-name} {token-path}

For example, on site-b:

$ cat <<EOF > token
eyJmc2lkIjoiOWY1MjgyZGItYjg5OS00NTk2LTgwOTgtMzIwYzFmYzM5NmYzIiwiY2xpZW50X2lkIjoicmJkLW1pcnJvci1wZWVyIiwia2V5IjoiQVFBUnczOWQwdkhvQmhBQVlMM1I4RmR5dHNJQU50bkFTZ0lOTVE9PSIsIm1vbl9ob3N0IjoiW3YyOjE5Mi4xNjguMS4zOjY4MjAsdjE6MTkyLjE2OC4xLjM6NjgyMV0ifQ==
EOF
$ rbd --cluster site-b mirror pool peer bootstrap import --site-name site-b image-pool token

Add Cluster Peer Manually

Cluster peers can be specified manually if desired or if the above bootstrap commands are not available with the currently installed Ceph release.

The remote rbd-mirror daemon will need access to the local cluster to perform mirroring. A new local Ceph user should be created for the remote daemon to use. To create a Ceph user, with ceph specify the auth get-or-create command, user name, monitor caps, and OSD caps:

$ ceph auth get-or-create client.rbd-mirror-peer mon 'profile rbd-mirror-peer' osd 'profile rbd'

The resulting keyring should be copied to the other cluster’s rbd-mirror daemon hosts if not using the Ceph monitor config-key store described below.

To manually add a mirroring peer Ceph cluster with rbd, specify the mirror pool peer add command, the pool name, and a cluster specification:

rbd mirror pool peer add {pool-name} {client-name}@{cluster-name}

For example:

$ rbd --cluster site-a mirror pool peer add image-pool client.rbd-mirror-peer@site-b
$ rbd --cluster site-b mirror pool peer add image-pool client.rbd-mirror-peer@site-a

By default, the rbd-mirror daemon needs to have access to a Ceph configuration file located at /etc/ceph/{cluster-name}.conf that provides the addresses of the peer cluster’s monitors, in addition to a keyring for {client-name} located in the default or configured keyring search paths (e.g. /etc/ceph/{cluster-name}.{client-name}.keyring).

Alternatively, the peer cluster’s monitor and/or client key can be securely stored within the local Ceph monitor config-key store. To specify the peer cluster connection attributes when adding a mirroring peer, use the --remote-mon-host and --remote-key-file optionals. For example:

$ cat <<EOF > remote-key-file
AQAeuZdbMMoBChAAcj++/XUxNOLFaWdtTREEsw==
EOF
$ rbd --cluster site-a mirror pool peer add image-pool client.rbd-mirror-peer@site-b --remote-mon-host 192.168.1.1,192.168.1.2 --remote-key-file remote-key-file
$ rbd --cluster site-a mirror pool info image-pool --all
Mode: pool
Peers:
  UUID                                 NAME   CLIENT                 MON_HOST                KEY
  587b08db-3d33-4f32-8af8-421e77abb081 site-b client.rbd-mirror-peer 192.168.1.1,192.168.1.2 AQAeuZdbMMoBChAAcj++/XUxNOLFaWdtTREEsw==

Remove Cluster Peer

To remove a mirroring peer Ceph cluster with rbd, specify the mirror pool peer remove command, the pool name, and the peer UUID (available from the rbd mirror pool info command):

rbd mirror pool peer remove {pool-name} {peer-uuid}

For example:

$ rbd --cluster site-a mirror pool peer remove image-pool 55672766-c02b-4729-8567-f13a66893445
$ rbd --cluster site-b mirror pool peer remove image-pool 60c0e299-b38f-4234-91f6-eed0a367be08

Data Pools

When creating images in the destination cluster, rbd-mirror selects a data pool as follows:

  1. If the destination cluster has a default data pool configured (with the rbd_default_data_pool configuration option), it will be used.

  2. Otherwise, if the source image uses a separate data pool, and a pool with the same name exists on the destination cluster, that pool will be used.

  3. If neither of the above is true, no data pool will be set.

Image Configuration

Unlike pool configuration, image configuration only needs to be performed against a single mirroring peer Ceph cluster.

Mirrored RBD images are designated as either primary or non-primary. This is a property of the image and not the pool. Images that are designated as non-primary cannot be modified.

Images are automatically promoted to primary when mirroring is first enabled on an image (either implicitly if the pool mirror mode was pool and the image has the journaling image feature enabled, or explicitly enabled by the rbd command if the pool mirror mode was image).

Enable Image Mirroring

If mirroring is configured in image mode for the image’s pool, then it is necessary to explicitly enable mirroring for each image within the pool. To enable mirroring for a specific image with rbd, specify the mirror image enable command along with the pool, image name, and mode:

rbd mirror image enable {pool-name}/{image-name} {mode}

The mirror image mode can either be journal or snapshot:

  • journal (default): When configured in journal mode, mirroring will utilize the RBD journaling image feature to replicate the image contents. If the RBD journaling image feature is not yet enabled on the image, it will be automatically enabled.

  • snapshot: When configured in snapshot mode, mirroring will utilize RBD image mirror-snapshots to replicate the image contents. Once enabled, an initial mirror-snapshot will automatically be created. Additional RBD image mirror-snapshots can be created by the rbd command.

For example:

$ rbd --cluster site-a mirror image enable image-pool/image-1 snapshot
$ rbd --cluster site-a mirror image enable image-pool/image-2 journal

Enable Image Journaling Feature

RBD journal-based mirroring uses the RBD image journaling feature to ensure that the replicated image always remains crash-consistent. When using the image mirroring mode, the journaling feature will be automatically enabled when mirroring is enabled on the image. When using the pool mirroring mode, before an image can be mirrored to a peer cluster, the RBD image journaling feature must be enabled. The feature can be enabled at image creation time by providing the --image-feature exclusive-lock,journaling option to the rbd command.

Alternatively, the journaling feature can be dynamically enabled on pre-existing RBD images. To enable journaling with rbd, specify the feature enable command, the pool and image name, and the feature name:

rbd feature enable {pool-name}/{image-name} {feature-name}

For example:

$ rbd --cluster site-a feature enable image-pool/image-1 journaling

Note

The journaling feature is dependent on the exclusive-lock feature. If the exclusive-lock feature is not already enabled, it should be enabled prior to enabling the journaling feature.

Tip

You can enable journaling on all new images by default by adding rbd default features = 125 to your Ceph configuration file.

Tip

rbd-mirror tunables are set by default to values suitable for mirroring an entire pool. When using rbd-mirror to migrate single volumes been clusters you may achieve substantial performance gains by setting rbd_mirror_journal_max_fetch_bytes=33554432 and rbd_journal_max_payload_bytes=8388608 within the [client] config section of the local or centralized configuration. Note that these settings may allow rbd-mirror to present a substantial write workload to the destination cluster: monitor cluster performance closely during migrations and test carefully before running multiple migrations in parallel.

Create Image Mirror-Snapshots

When using snapshot-based mirroring, mirror-snapshots will need to be created whenever it is desired to mirror the changed contents of the RBD image. To create a mirror-snapshot manually with rbd, specify the mirror image snapshot command along with the pool and image name:

rbd mirror image snapshot {pool-name}/{image-name}

For example:

$ rbd --cluster site-a mirror image snapshot image-pool/image-1

By default up to 5 mirror-snapshots will be created per-image. The most recent mirror-snapshot is automatically pruned if the limit is reached. The limit can be overridden via the rbd_mirroring_max_mirroring_snapshots configuration option if required. Additionally, mirror-snapshots are automatically deleted when the image is removed or when mirroring is disabled.

Mirror-snapshots can also be automatically created on a periodic basis if mirror-snapshot schedules are defined. The mirror-snapshot can be scheduled globally, per-pool, or per-image levels. Multiple mirror-snapshot schedules can be defined at any level, but only the most-specific snapshot schedules that match an individual mirrored image will run.

To create a mirror-snapshot schedule with rbd, specify the mirror snapshot schedule add command along with an optional pool or image name; interval; and optional start time:

rbd mirror snapshot schedule add [--pool {pool-name}] [--image {image-name}] {interval} [{start-time}]

The interval can be specified in days, hours, or minutes using d, h, m suffix respectively. The optional start-time can be specified using the ISO 8601 time format. For example:

$ rbd --cluster site-a mirror snapshot schedule add --pool image-pool 24h 14:00:00-05:00
$ rbd --cluster site-a mirror snapshot schedule add --pool image-pool --image image1 6h

To remove a mirror-snapshot schedules with rbd, specify the mirror snapshot schedule remove command with options that match the corresponding add schedule command.

To list all snapshot schedules for a specific level (global, pool, or image) with rbd, specify the mirror snapshot schedule ls command along with an optional pool or image name. Additionally, the --recursive option can be specified to list all schedules at the specified level and below. For example:

$ rbd --cluster site-a mirror snapshot schedule ls --pool image-pool --recursive
POOL        NAMESPACE IMAGE  SCHEDULE
image-pool  -         -      every 1d starting at 14:00:00-05:00
image-pool            image1 every 6h

To view the status for when the next snapshots will be created for snapshot-based mirroring RBD images with rbd, specify the mirror snapshot schedule status command along with an optional pool or image name:

rbd mirror snapshot schedule status [--pool {pool-name}] [--image {image-name}]

For example:

$ rbd --cluster site-a mirror snapshot schedule status
SCHEDULE TIME       IMAGE
2020-02-26 18:00:00 image-pool/image1

Disable Image Mirroring

To disable mirroring for a specific image with rbd, specify the mirror image disable command along with the pool and image name:

rbd mirror image disable {pool-name}/{image-name}

For example:

$ rbd --cluster site-a mirror image disable image-pool/image-1

Image Promotion and Demotion

In a failover scenario where the primary designation needs to be moved to the image in the peer Ceph cluster, access to the primary image should be stopped (e.g. power down the VM or remove the associated drive from a VM), demote the current primary image, promote the new primary image, and resume access to the image on the alternate cluster.

Note

RBD only provides the necessary tools to facilitate an orderly failover of an image. An external mechanism is required to coordinate the full failover process (e.g. closing the image before demotion).

To demote a specific image to non-primary with rbd, specify the mirror image demote command along with the pool and image name:

rbd mirror image demote {pool-name}/{image-name}

For example:

$ rbd --cluster site-a mirror image demote image-pool/image-1

To demote all primary images within a pool to non-primary with rbd, specify the mirror pool demote command along with the pool name:

rbd mirror pool demote {pool-name}

For example:

$ rbd --cluster site-a mirror pool demote image-pool

To promote a specific image to primary with rbd, specify the mirror image promote command along with the pool and image name:

rbd mirror image promote [--force] {pool-name}/{image-name}

For example:

$ rbd --cluster site-b mirror image promote image-pool/image-1

To promote all non-primary images within a pool to primary with rbd, specify the mirror pool promote command along with the pool name:

rbd mirror pool promote [--force] {pool-name}

For example:

$ rbd --cluster site-a mirror pool promote image-pool

Tip

Since the primary / non-primary status is per-image, it is possible to have two clusters split the IO load and stage failover / failback.

Note

Promotion can be forced using the --force option. Forced promotion is needed when the demotion cannot be propagated to the peer Ceph cluster (e.g. Ceph cluster failure, communication outage). This will result in a split-brain scenario between the two peers and the image will no longer be in-sync until a force resync command is issued.

Force Image Resync

If a split-brain event is detected by the rbd-mirror daemon, it will not attempt to mirror the affected image until corrected. To resume mirroring for an image, first demote the image determined to be out-of-date and then request a resync to the primary image. To request an image resync with rbd, specify the mirror image resync command along with the pool and image name:

rbd mirror image resync {pool-name}/{image-name}

For example:

$ rbd mirror image resync image-pool/image-1

Note

The rbd command only flags the image as requiring a resync. The local cluster’s rbd-mirror daemon process is responsible for performing the resync asynchronously.

Mirror Status

The peer cluster replication status is stored for every primary mirrored image. This status can be retrieved using the mirror image status and mirror pool status commands.

To request the mirror image status with rbd, specify the mirror image status command along with the pool and image name:

rbd mirror image status {pool-name}/{image-name}

For example:

$ rbd mirror image status image-pool/image-1

To request the mirror pool summary status with rbd, specify the mirror pool status command along with the pool name:

rbd mirror pool status {pool-name}

For example:

$ rbd mirror pool status image-pool

Note

Adding --verbose option to the mirror pool status command will additionally output status details for every mirroring image in the pool.

rbd-mirror Daemon

The two rbd-mirror daemons are responsible for watching image journals on the remote, peer cluster and replaying the journal events against the local cluster. The RBD image journaling feature records all modifications to the image in the order they occur. This ensures that a crash-consistent mirror of the remote image is available locally.

The rbd-mirror daemon is available within the optional rbd-mirror distribution package.

Important

Each rbd-mirror daemon requires the ability to connect to both clusters simultaneously.

Warning

Pre-Luminous releases: only run a single rbd-mirror daemon per Ceph cluster.

Each rbd-mirror daemon should use a unique Ceph user ID. To create a Ceph user, with ceph specify the auth get-or-create command, user name, monitor caps, and OSD caps:

ceph auth get-or-create client.rbd-mirror.{unique id} mon 'profile rbd-mirror' osd 'profile rbd'

The rbd-mirror daemon can be managed by systemd by specifying the user ID as the daemon instance:

systemctl enable ceph-rbd-mirror@rbd-mirror.{unique id}

The rbd-mirror can also be run in foreground by rbd-mirror command:

rbd-mirror -f --log-file={log_path}