Creating a Volume
Describes how to create a volume using the Control System, CLI and the REST API.
About this task
You can create a new (Standard or Mirror) volume using the Control System, the CLI, and the REST API.
Creating a Volume Using the Control System
About this task
To create a new (Standard or Mirror) volume using the Control System:
Procedure
-
Go to the Create Volume to
display the Create New Volume page.
page and click NOTEWhen running on a Kubernetes cluster, the Create Volume option is on the Volumes page.
-
Choose the Volume Type in the
Properties section. Choose:
- Standard Volume to create a read-write volume.
- Mirror Volume to create a volume that is a
read-only copy of an existing volume.TIPSee also: Mirror Types.
-
Specify the following required settings in the
Properties section:
Volume Name Enter a name for the volume. The name should contain only the following characters:A-Z a-z 0-9 _ - .
NOTE- The volume name should not begin with
mapr.
becausemapr.
is used for system volumes. If you usemapr.
at the start of the volume name, the volume may not display in the default view of the list of volumes in the Control System; you must select the Include System Volumes checkbox in the Volumes pane to view volumes with names beginning withmapr
. - For tiering-enabled volumes, volume name should not exceed ninety-eight characters.
Accountable Entity Specifies a user or group whose use of a volume can be subject to quotas. You can set or modify quotas that limit the space used by all the volumes owned by an accountable entity. Volume Name Enter a name for the volume. The name should contain only the following characters:A-Z a-z 0-9 _ - .
NOTE- The volume name should not begin with
mapr.
becausemapr.
is used for system volumes. If you usemapr.
at the start of the volume name, the volume may not display in the default view of the list of volumes in the Control System; you must select the Include System Volumes checkbox in the Volumes pane to view volumes with names beginning withmapr
. - For tiering-enabled volumes, volume name should not exceed ninety-eight characters.
Source Cluster Name Enter the name of the cluster on which the source volume exists. The name should contain only the following characters:Mirroring only works between two secure clusters or between two non-secure clusters. Mirroring does not work when one cluster is secure and the other is non-secure.A-Z a-z 0-9 _ - .
NOTEWhen setting up mirror volumes for mirroring between clusters, for the mirroring operation to run successfully, servers in one cluster cannot use the same IP addresses as servers in the other cluster. For example, if node A in cluster A has a private IP address of 10.10.20.29, no server in cluster B can have the same private IP address. Also, all the servers in destination cluster must be able to reach all the servers in the source cluster and vice versa. For example, suppose 10.10.20.29 is the only IP address used by node A in cluster A; then all servers in cluster B should be able to reach the IP address 10.10.20.29.Source Volume Name Enter the name of the source volume, from which the mirror volume pulls data (after selecting the source volume cluster). The name should contain only the following characters:If the source volume is on:A-Z a-z 0-9 _ - .
- Same cluster, you create a local mirror volume, which is useful for load balancing or for providing a read-only copy of a data set. See Local Mirroring for more information.
- Another cluster, you create a remote mirror volume, which is useful for offsite backup, for data transfer to remote facilities, and for load and latency balancing. See Creating Remote Mirrors for more information.
For information on setting up mirror cascades, see Mirror Cascades.NOTEIf you plan to enable tiering for the mirror volume, ensure that the selected source volume is also tiering-enabled. You cannot create tiering-enabled mirror volumes to mirror data in standard volumes that are not tiering-enabled.Accountable Entity Specifies a user or group whose use of a volume can be subject to quotas. You can set or modify quotas that limit the space used by all the volumes owned by an accountable entity. - The volume name should not begin with
- (Optional) Step 9 to associate a snapshot schedule and/or an offload schedule with the volume.
- Step 11 to create the volume with basic settings.
-
(Optional) Specify the following general settings under
Properties:
Mount Specifies whether to automatically mount (Yes) or not mount (No) the volume after creation. By default, volumes are mounted immediately after creation. If this is set to Yes, you must also specify the mount path. Mount Path The path to mount the volume. This is required if the value for Mount is Yes. NOTEThe path must be relative to/
and cannot be in the form of a global namespace path (for example,/mapr/<cluster-name>/
).Collect Metrics Specifies whether (Yes) or not (No) to enable metrics collection for this volume. For more information, see Collecting Volume Metrics and Enabling Volume Metric Collection. Volume Access Specifies whether the volume is read-only or a read/write volume. By default, a standard volume is created with read/write access. A mirror volume can only be a read-only volume. Last Access Interval Denotes the frequency at which the access time of a file is updated. See Tuning Last Access Time for more information. -
(Optional) Specify the following settings for data replication under
Replication and Storage section:
Topology Specifies the rack path to the volume. The default topology is /data
.Optimize Replication for Specifies the basis for the replication factor: - High throughput, or cascading replication, where volumes are replicated sequentially on intermediate and tail containers.
- Low latency, or star replication, where volumes are replicated on multiple containers in parallel.
Guarantee Min Replication Specifies whether (Yes) or not (No) to enforce minimum number of copies. If this is enabled (Yes), writes succeed only when the minimum number of copies exist. If this is enabled (Yes) and minimum number of copies are not available, the client is asked to retry. For more information, see Understanding Replication.
Replication Specifies the minimum (Minimum Replication) and desired (Target Replication) number of copies of the volume data. The default minimum is 2, and the default target is 3. Name Container Replication Specifies the minimum (Minimum Replication) and desired (Target Replication) number of copies of the name container associated with the volume. The default minimum is 2, and the default target is 3. Data Tiering — Specifies whether to enable (Yes) or disable (No) data tiering for volume data.By default, data tiering is enabled and volume data is stored in the hot tier (HPE Ezmeral Data Fabric cluster). If you choose to enable data tiering for the volume, you can associate a tier type with the volume either now, or later by editing the volume. If you decide to associate a type of tier with the volume, proceed to the next step; otherwise, proceed to step 7.
-
(Optional) Associate a type of tier with the volume by selecting a tiering
type from the Tiering Type drop-down list and
specifying the following settings for the tier:
For offloading data to an erasure coded volume, specify values for the following properties. If values are not specified, default values are applied.
Topology The topology of the erasure coded volume from the drop-down list. Storage Policy The rule for offloading data in this volume. You can click: - Browse to select an existing rule.
- Create to create a new rule for offloading data. See steps 3 - 5 in the Creating a Storage Tier Policy topic for more information.
default.ectier.rule
, which is all files (p
), is associated with the volume.Scheme The erasure coding scheme, which is the number of data chunks and number of parity chunks. Set the required Parity Scheme. The system indicates in real-time whether or not the parity scheme is valid. Some valid parity schemes include: - 3+2 — for 3 data chunks and 2 parity chunks. You must have 5 or more nodes on the cluster for this option. If selected, this scheme has 60% storage overhead and can tolerate failure of up to 2 nodes.
- 4+2 — for 4 data chunks and 2 parity chunks. You must have 6 or more nodes on the cluster for this option. If selected, this scheme has 50% storage overhead and can tolerate failure of up to 2 nodes.
- 5+2 — for 5 data chunks and 2 parity chunks. You must have 7 or more nodes on the cluster for this option. If selected, this scheme has 40% storage overhead and can tolerate failure of up to 2 nodes.
- 6+3 — for 6 data chunks and 3 parity chunks. You must have 9 or more nodes on the cluster for this option. If selected, this scheme has 50% storage overhead and can tolerate failure of up to 3 nodes.
Yes
. The system then displays a third slider to set the number of local parity blocks.As you set the parity scheme, irrespective of local or not, the system indicates the number of failures that the parity scheme can tolerate, the storage overhead required, and the number of nodes required to implement the parity scheme.
See Erasure Coding Scheme for Data Protection and Recovery for more information.NOTEAlthough you can create a volume even if the required number of nodes are not present, offload operation fails if the required number of nodes are not present.For offloading data to a low cost storage alternative on the cloud, specify values for the following properties.Storage Policy The rule for offloading data in this volume. You can click: - Browse to select an existing rule.
- Create to create a new rule for offloading data. See Creating a Storage Tier Policy for more information.
Remote Target The location to which the data is offloaded. You can click: - Browse to select an existing tier.
- Create to create a new tier. See Creating a Storage Tier for more information.
Retention Duration after Recall The number of days to retain data recalled from the tier to the HPE Ezmeral Data Fabric cluster. Once the number of days is reached, recalled data on the HPE Ezmeral Data Fabric cluster is purged (if there are no changes), or offloaded (if there are changes). Tier Encryption Specifies whether (Yes) or not (No) to enable encryption of data on the tier. This cannot be modified once it is set. The default value is No (disabled). - (Optional) To use Label-Based Storage, enter the label and the namespace label . See Using Storage Labels for more information.
-
(Optional) Configure security for volume data by setting values for
the properties in the Security section.
-
(Optional) Specify the users, groups, and/or roles that have and/or do not
have permissions to read and/or write in the Data Access
Control section:
Click:
- to modify the users, groups, and/or roles.
- Add Another to grant permissions for other users, groups, and/or roles and repeat steps b and c.
- to create a copy of the permissions, which you can then modify.
- to remove a data access control setting.
-
(Optional) Specify the users and groups that have volume administration
permissions:
To define administrative access control settings for another user or group, click Add Another and repeat steps a and b.By default, the root user and the volume creator have full control permissions on the volume.NOTETo perform this action from the command line, refer to acl set.
- Set read (R), write (W), and/or execute (X) permissions on the root directory for users, groups, and others by selecting the permission.
- Click Create Volume to create the volume.
Creating a Volume Using the CLI and REST API
About this task
/opt/mapr/bin/maprcli volume create -name <volName> -path <mountPath>
A-Z a-z 0-9 _ - .
- Mirror volume, you must specify
-type mirror
and-source <sourceVolName>@<cluster>
in the command. - Tiering-enabled volume, you must specify
-tieringenable true
in the command.
curl -X POST 'https://<hostname>:8443/rest/volume/create?name=<volName>&path=<mountPath>' --user <username>:<password>
The name should contain only the following characters:
A-Z a-z 0-9 _ - .
If you are creating a:- Mirror volume, you must specify
type=mirror
andsource=<sourceVolName>@<cluster>
in the request. - Tiering-enabled volume, you must specify
tieringenable=true
in the request. Thetieringenable
property of a mirror volume should be the same as the source volume.
volume create
.