Replicating Block Storage for Classic Volumes
Replication uses one of your snapshot schedules to automatically copy snapshots to a destination volume in a remote data centers. The copies can be recovered in the remote site if a catastrophic event occurs or your data becomes corrupted.
Replication keeps your data in sync in two different locations. If you want to clone your volume and use it independently from the original volume, see Creating a duplicate Block Volume.
Before you can replicate, you must create a snapshot schedule. The option to Order Replica does not appear unless this condition is met.
Determining the remote data center for my replicated storage volume in the console
IBM Cloud data centers are paired into primary and remote combinations in every region worldwide. When you replicate data, consider the local data residency laws because moving data across borders can have legal implications. Replication across regions is not permitted.
The following table shows the data center codes within each region.
US | Latin America | Canada | Europe | Asia-Pacific | Australia |
---|---|---|---|---|---|
|
|
|
|
|
|
Determining the remote location for the replicated storage volume from the CLI
Before you begin, decide on the CLI client that you want to use.
- You can either install the IBM Cloud CLI and install the SL plug-in with
ibmcloud plugin install sl
. For more information, see Extending IBM Cloud CLI with plug-ins. - Or, you can install the SLCLI.
IBM Cloud data centers are paired into primary and remote combinations in every region worldwide. When you replicate data, consider the local data residency laws because moving data across borders can have legal implications. Replication across regions is not permitted.
Listing locations from the IBMCLOUD CLI
You can use the ibmcloud sl block replica-locations
command to locate a replica location for your block volume. The following example lists the available location for a block volume that is located in the US south region.
$ ibmcloud sl block replica-locations 560156918
ID Short Name Long Name
449494 dal09 Dallas 9
957095 wdc04 Washington 4
1004995 sjc03 San Jose 3
1441195 dal10 Dallas 10
1854795 dal12 Dallas 12
2017603 wdc07 Washington 7
2017695 wdc06 Washington 6
2178495 sjc04 San Jose 4
For more information about all of the parameters that are available for this command, see ibmcloud sl block replica-locations.
Listing locations from the SLCLI
To list suitable replication data centers for a specific volume, use the following command.
$ slcli block replica-locations --help
Usage: slcli block replica-locations [OPTIONS] VOLUME_ID
Options:
--sortby TEXT Column to sort by
--columns TEXT Columns to display. Options: ID, Long Name, Short Name
-h, --help Show this message and exit.
As part of the data center modernization strategy for IBM Cloud®, several data centers are scheduled to consolidate in 2023. For more information, see Data center consolidations.
Creating the initial replica in the console
Replications work based on a snapshot schedule. You must first have snapshot space and a snapshot schedule for the source volume before you can replicate. If you try to configure replication and one or the other isn't in place, you are going to be prompted to purchase more space or set up a schedule. Replications are managed under Storage, Block Storage for Classic in the IBM Cloud® console.
-
Click the name of your storage volume to display its details.
-
Click Actions and click Order Replica.
-
Select the existing snapshot schedule that you want your replication to follow. The list contains all of your active snapshot schedules.
You can select only one schedule even if you have a mix of hourly, daily, and weekly. All snapshots that were captured after the previous replication are copied over to the replica. For more information, see Working with Snapshots. Replication starts 5 minutes after the snapshot is taken to ensure that the most up-to-date data is copied to the replica volume.
-
Select a Location for the replica volume.
-
Click Continue.
-
Enter in a Promo Code if you have one, and click Recalculate. The other fields in the window are completed by default.
Discounts are applied when the order is processed.
-
Review your order, and read the service agreement. If you agree with the terms, check the box.
-
Click Place Order.
Creating the initial replica from the CLI
Replications work based on a snapshot schedule. You must first have snapshot space and a snapshot schedule for the source volume before you can replicate.
Creating the replica volume from the IBMCLOUD CLI
You can use the ibmcloud sl block replica-order
command to order a replica volume. Specify the source volume, the replication schedule, and the location of your replica volume.
$ ibmcloud sl block replica-order 562193766 --snapshot-schedule DAILY -d dal13 -t 0.25 -o LINUX
This action will incur charges on your account. Continue?> y
For more information about all of the parameters that are available for this command, see ibmcloud sl block replica-order.
Creating the replica volume from the SLCLI
You can use the following command to order a replica volume.
$ slcli block replica-order --help
Usage: slcli block replica-order [OPTIONS] VOLUME_ID
Options:
-s, --snapshot-schedule [INTERVAL|HOURLY|DAILY|WEEKLY]
Snapshot schedule to use for replication,
(INTERVAL | HOURLY | DAILY | WEEKLY)
[required]
-l, --location TEXT Short name of the data center for the
replicant (e.g.: dal09) [required]
--tier [0.25|2|4|10] Endurance Storage Tier (IOPS per GB) of the
primary volume for which a replicant is
ordered [optional]
-h, --help Show this message and exit.
Viewing the replica volumes in the Volume List in the console
You can view your replication volumes on the Block Storage for Classic page under Storage > Block Storage for Classic. The Volume Name shows the primary volume's name followed by REP. The Type is Endurance or Performance – Replica.
Viewing the replica volumes from the CLI
Listing replica volumes from the IBMCLOUD CLI
You can use the ibmcloud sl block replica-order
command to list the replicas of your block volume. The following example lists the replica partners of the block volume 560156918
.
$ ibmcloud sl block replica-partners 560156918
ID User name Account ID Capacity (GB) Hardware ID Guest ID Host ID
560382016 SL02SEV1414935_268_REP_1 1234567 500 - - -
For more information about all of the parameters that are available for this command, see ibmcloud sl block replica-partners.
Listing replica volumes from the SLCLI
List existing replicant volumes for a block volume with the following command.
$ slcli block replica-partners --help
Usage: slcli block replica-partners [OPTIONS] VOLUME_ID
Options:
--sortby TEXT Column to sort by
--columns TEXT Columns to display. Options: ID, Username, Account ID,
Capacity (GB), Hardware ID, Guest ID, Host ID
-h, --help Show this message and exit.
Viewing the replication history in the console
To view the Replication history, click Manage on the main menu bar. Select Account, and scroll to the Audit Log. The Storage Replication Events list contains the names of the volume, a description of the replication event and the timestamp of the event.
Editing the Replication Schedule in the console
The replication schedule is based on an existing snapshot schedule. To change the replica schedule from Hourly to Daily or Weekly or vice versa, you must cancel the replica volume and set up a new one.
However, if you want to change the time of day when your Daily replication occurs, you can adjust the existing schedule on the active volume.
- On the active volume details page, click Actions .
- Select Edit Snapshot Schedule.
- Look in the Snapshot frame under Schedule to determine which Daily schedule you're using for replication. Change the schedule that you want.
- Click Save.
Deleting an existing replica block volume in the console
You can cancel replication either immediately or on the anniversary date, which causes billing to end.
- Click the volume from the Block Storage for Classic page.
- Click Actions .
- Select Delete Replica.
- Select when to cancel. Choose Immediately or Anniversary Date, and click Continue.
- This operation deletes the replica volume with all its data. Click the checkbox to acknowledge the information, and click Delete.
Canceling replication when the primary volume is deleted in the console
When a primary volume is canceled, the replication schedule and the volume in the replica data center are deleted. Replicas are canceled from the Block Storage for Classic page.
You can expect the volume to remain visible in your Storage list for at least 24 hours (immediate cancellation) or until the anniversary date. Certain features aren't going to be available any longer, but the volume remains visible until it is reclaimed. However, billing is stopped immediately after you click Delete Replica.
Active replicas can block reclamation of the Storage volume. Make sure that the volume is no longer mounted, host authorizations are revoked, and replication is canceled before you attempt to cancel the original volume.
Creating a duplicate of a replica
You can create a duplicate of an existing IBM Cloud® Block Storage for Classic. The duplicate volume inherits the capacity and performance options of the original volume by default and has a copy of the data up to the point-in-time of a snapshot.
Duplicates can be created from both primary and replica volumes. The new duplicate is created in the same data center as the original volume. If you create a duplicate from a replica volume, the new volume is created in the same data center as the replica volume.
Duplicate volumes can be accessed by a host for read/write as soon as the storage is provisioned. However, snapshots and replication aren't allowed until the data copy from the original to the duplicate is complete.
For more information, see Creating a duplicate Block Storage for Classic.
Using replicas to failover when disaster strikes
When you fail over, you’re "flipping the switch" from your storage volume in your primary data center to the destination volume in your remote data center. For example, your primary data center is London and your secondary data center is Amsterdam. If a failure event occurs, you’d fail over to Amsterdam – connecting to the now-primary volume from a Compute instance in Amsterdam. After your volume in London is repaired, a snapshot is taken of the Amsterdam volume to fail back to London and the once-again primary volume from a Compute instance in London.
- If the primary location is in imminent danger or severely impacted, see Failover with an accessible primary volume.
- If the primary location is down, see Failover with an inaccessible primary volume.