Kaminario K2 all-flash array iSCSI and FC volume drivers¶

Driver requirements¶

• Kaminario’s K2 all-flash iSCSI and/or FC array

• K2 REST API version >= 2.2.0

• K2 version 5.8 or later are supported

• krest python library(version 1.3.1 or later) should be installed on the Block Storage node using sudo pip install krest

• The Block Storage Node should also have a data path to the K2 array for the following operations:

  • Create a volume from snapshot

  • Clone a volume

  • Copy volume to image

  • Copy image to volume

  • Retype ‘dedup without replication’<->’nodedup without replication’

Supported operations¶

• Create, delete, attach, and detach volumes.

• Create and delete volume snapshots.

• Create a volume from a snapshot.

• Copy an image to a volume.

• Copy a volume to an image.

• Clone a volume.

• Extend a volume.

• Retype a volume.

• Manage and unmanage a volume.

• Replicate volume with failover and failback support to K2 array.

Limitations and known issues¶

If your OpenStack deployment is not setup to use multipath, the network connectivity of the K2 all-flash array will use a single physical port.

This may significantly limit the following benefits provided by K2:

• available bandwidth

• high-availability

• non disruptive-upgrade

The following steps are required to setup multipath access on the Compute and the Block Storage nodes

1. Install multipath software on both Compute and Block Storage nodes.

   For example:

   # apt-get install sg3-utils multipath-tools
   

2. In the [libvirt] section of the nova.conf configuration file, specify volume_use_multipath=True. This option is valid for both iSCSI and FC drivers. In versions prior to Newton, the option was called iscsi_use_multipath.

   Additional resources: Kaminario Host Configuration Guide for Linux (for configuring multipath)

3. Restart the compute service for the changes to take effect.

   # service nova-compute restart
   

Configure single Kaminario iSCSI/FC back end¶

This section details the steps required to configure the Kaminario Cinder Driver for single FC or iSCSI backend.

1. In the cinder.conf configuration file under the [DEFAULT] section, set the scheduler_default_filters parameter:

   [DEFAULT]
   scheduler_default_filters = DriverFilter,CapabilitiesFilter
   

   See following documents for more information: Cinder Scheduler Filters and Configure and use driver filter and weighing for scheduler.

2. Under the [DEFAULT] section, set the enabled_backends parameter with the iSCSI or FC back-end group

   [DEFAULT]
   # For iSCSI
   enabled_backends = kaminario-iscsi-1
   
   # For FC
   # enabled_backends = kaminario-fc-1
   

3. Add a back-end group section for back-end group specified in the enabled_backends parameter

4. In the newly created back-end group section, set the following configuration options:

   [kaminario-iscsi-1]
   # Management IP of Kaminario K2 All-Flash iSCSI/FC array
   san_ip = 10.0.0.10
   # Management username of Kaminario K2 All-Flash iSCSI/FC array
   san_login = username
   # Management password of Kaminario K2 All-Flash iSCSI/FC array
   san_password = password
   # Enable Kaminario K2 iSCSI/FC driver
   volume_driver = cinder.volume.drivers.kaminario.kaminario_iscsi.KaminarioISCSIDriver
   # volume_driver = cinder.volume.drivers.kaminario.kaminario_fc.KaminarioFCDriver
   
   # Backend name
   # volume_backend_name = kaminario_fc_1
   volume_backend_name = kaminario_iscsi_1
   
   # K2 driver calculates max_oversubscription_ratio on setting below
   # option as True. Default value is False
   # auto_calc_max_oversubscription_ratio = False
   
   # Set a limit on total number of volumes to be created on K2 array, for example:
   # filter_function = "capabilities.total_volumes < 250"
   
   # For replication, replication_device must be set and the replication peer must be configured
   # on the primary and the secondary K2 arrays
   # Syntax:
   #     replication_device = backend_id:<s-array-ip>,login:<s-username>,password:<s-password>,rpo:<value>
   # where:
   #     s-array-ip is the secondary K2 array IP
   #     rpo must be either 60(1 min) or multiple of 300(5 min)
   # Example:
   # replication_device = backend_id:10.0.0.50,login:kaminario,password:kaminario,rpo:300
   
   # Suppress requests library SSL certificate warnings on setting this option as True
   # Default value is 'False'
   # suppress_requests_ssl_warnings = False
   

5. Restart the Block Storage services for the changes to take effect:

   # service cinder-api restart
   # service cinder-scheduler restart
   # service cinder-volume restart
   

Setting multiple Kaminario iSCSI/FC back ends¶

The following steps are required to configure multiple K2 iSCSI/FC backends:

1. In the cinder.conf file under the [DEFAULT] section, set the enabled_backends parameter with the comma-separated iSCSI/FC back-end groups.

   [DEFAULT]
   enabled_backends = kaminario-iscsi-1, kaminario-iscsi-2, kaminario-iscsi-3
   

2. Add a back-end group section for each back-end group specified in the enabled_backends parameter

3. For each back-end group section, enter the configuration options as described in the above section Configure single Kaminario iSCSI/FC back end

   See Configure multiple-storage back ends for additional information.

4. Restart the cinder volume service for the changes to take effect.

   # service cinder-volume restart
   

Creating volume types¶

Create volume types for supporting volume creation on the multiple K2 iSCSI/FC backends. Set following extras-specs in the volume types:

• volume_backend_name : Set value of this spec according to the value of volume_backend_name in the back-end group sections. If only this spec is set, then dedup Kaminario cinder volumes will be created without replication support

  $ openstack volume type create kaminario_iscsi_dedup_noreplication
  $ openstack volume type set --property volume_backend_name=kaminario_iscsi_1 \
    kaminario_iscsi_dedup_noreplication
  

• kaminario:thin_prov_type : Set this spec in the volume type for creating nodedup Kaminario cinder volumes. If this spec is not set, dedup Kaminario cinder volumes will be created.

• kaminario:replication : Set this spec in the volume type for creating replication supported Kaminario cinder volumes. If this spec is not set, then Kaminario cinder volumes will be created without replication support.

  $ openstack volume type create kaminario_iscsi_dedup_replication
  $ openstack volume type set --property volume_backend_name=kaminario_iscsi_1 \
    kaminario:replication=enabled kaminario_iscsi_dedup_replication
  
  $ openstack volume type create kaminario_iscsi_nodedup_replication
  $ openstack volume type set --property volume_backend_name=kaminario_iscsi_1 \
    kaminario:replication=enabled kaminario:thin_prov_type=nodedup \
    kaminario_iscsi_nodedup_replication
  
  $ openstack volume type create kaminario_iscsi_nodedup_noreplication
  $ openstack volume type set --property volume_backend_name=kaminario_iscsi_1 \
    kaminario:thin_prov_type=nodedup kaminario_iscsi_nodedup_noreplication
  

Supported retype cases¶

The following are the supported retypes for Kaminario cinder volumes:

• Nodedup-noreplication <–> Nodedup-replication

  $ cinder retype volume-id new-type
  

• Dedup-noreplication <–> Dedup-replication

  $ cinder retype volume-id new-type
  

• Dedup-noreplication <–> Nodedup-noreplication

  $ cinder retype --migration-policy on-demand volume-id new-type
  

For non-supported cases, try combinations of the cinder retype command.

Driver options¶

The following table contains the configuration options that are specific to the Kaminario K2 FC and iSCSI Block Storage drivers.