Creating a IBM Storage Scale Primary configuration pattern
Use the Pattern Builder to create a virtual application pattern for deploying a IBM Storage Scale Server Primary configuration.
Before you begin
About this task
When
you create a IBM Storage Scale Server virtual
application, you select the GPFS Pattern Type 1.2
pattern
type. This pattern type abstracts the infrastructure and middleware
layers for the IBM Storage Scale workload.
You then create a virtual application pattern by
using the components that are associated with this pattern type.
Use the Pattern Builder to define, create, and deploy your virtual applications. For more information about creating virtual application patterns in Pattern Builder, see the Related links.
Procedure
- Click Patterns > Virtual Application Patterns.
- Click Create New.
- To build your virtual application pattern:
- Select the
GPFS Pattern Type 1.2
pattern type from the menu. - Select the
GPFS Primary
virtual application template. - Click Start Building. The Pattern Builder opens in a new window where you can add components and policies.
- Select the
- In the pattern properties pane, specify the following information:
- Name
- The name of the virtual application pattern, such as IBM Storage Scale Server - Primary Config.
- Description
- The description of the virtual application pattern. This field is optional.
- Type
- Leave Pattern selected to create a virtual application pattern or select Pattern Template to create a virtual application template that is used as the basis for creating other virtual application patterns.
- Lock option for plug-in usage
- Specify how this virtual application pattern is
affected by upgrades to the pattern type or to IBM® Foundation Pattern.
- Unlock plug-ins
- If the pattern type is upgraded, use the latest versions of pattern type plug-ins. If IBM Foundation Pattern is upgraded, use the latest version.
- Lock all plug-ins except Foundation plug-ins
- If the pattern type is upgraded, do not change the version of the plug-ins that are associated with this virtual application pattern. If IBM Foundation Pattern is upgraded, use the latest version.
- Lock all plug-ins
- Do not change the version of plug-ins or the version of the IBM Foundation Pattern associated with this virtual application pattern when an upgrade occurs.
Note: If you select Lock all plug-ins except Foundation plug-ins or Lock all plug-ins, you can view a list of which plug-ins are locked. Click the Source tab in Pattern Builder. The application model source is displayed. Search for the elementplugins
to view the list.
- Click Save and save the virtual
pattern definition. You can alternatively click Save As to save as a new virtual application template that you can use to create other virtual application patterns.
- Configure the IBM Storage Scale Server component
in Pattern Builder.
- With the IBM Storage Scale Server component
selected in the canvas, configure the following settings in the properties
pane:
- Name
- A unique name that identifies the IBM Storage Scale Server component in the Pattern Builder.
- Key Management section
- You can optionally provide the various SSH keys that the deployed IBM Storage Scale configuration uses. You can
also specify keys at deployment time. These keys are generated on
each deployment automatically if no keys are provided by the deployer.
You can retrieve keys that are generated or updated after deployment,
by using the Manage > Operations > GPFS_Manager operations view.
If you do specify any keys at deployment time, ensure that you specify
both private and public keys, using a tar.gz compression
format.You can specify the following types of keys:
- IBM Storage Scale Manager key
- This key is used by theIBM Storage Scale Manager virtual
machines to communicate with other IBM Storage Scale Manager virtual
machines.
For example, when you attach a IBM Storage Scale Mirror or IBM Storage Scale Tiebreaker configuration to a IBM Storage Scale Primary configuration, all three managers, from the Primary, Mirror, and Tiebreaker deployments must have the same IBM Storage Scale Manager key installed. If not, you receive an error message indicating that the IBM Storage Scale Manager virtual machine from the Primary configuration is unable to communicate with the IBM Storage Scale Manager from the Mirror or Tiebreaker configurations. You must then install the same IBM Storage Scale Manager key on all managers, and try to attach again.
When you create the IBM Storage Scale Server component by using the virtual application template, a default IBM Storage Scale Managers key, artifacts/manager_key_linux.tar.gz is provided for you in this field.
All IBM Storage Scale Server virtual application templates use the same default tar.gz file for the IBM Storage Scale Managers key, so when you deploy Primary, Mirror, or Tiebreaker configurations using these virtual application templates, their managers can already communicate with each other because they are sharing the same key.
If you need to install the same IBM Storage Scale Managers key in these various configurations, you can follow this procedure:- Retrieve the IBM Storage Scale Managers key from the Primary configuration. Select Manage > Operations > GPFS_Manager > Manage Keys > Retrieve Key and select Manager.
- Pass the retrieved tar.gz file to the Mirror,
or Tiebreak configurations by using either of the following methods:
- Specify the tar.gz file in the Key Management section of the configuration during deployment time.
- If the configurations are already deployed, select Manage > Operations > GPFS_Manager > Manage Keys > Install Key, then select Manager, and specify the tar.gz file obtained from the Primary configuration.
- IBM Storage Scale cluster key
- This key is installed on all of the IBM Storage Scale Server nodes, and is used by the IBM Storage Scale application to communicate within the nodes server cluster.
- IBM Storage Scale Clients key
- This key is used to access the IBM Storage Scale Manager from a IBM Storage Scale Client. This is required so that the IBM Storage Scale Client can send requests to the IBM Storage Scale Manager virtual machine and pass any extra parameters required to add itself to the IBM Storage Scale cluster.
- IBM Storage Scale Configurations section
- In this section you select the type of IBM Storage Scale Server configuration
to deploy. Available choices include:
- Primary Configuration
- Mirror Configuration
- Passive ConfigurationNote: This configuration type is not supported in Cloud Pak System Software for Power®.
- Tiebreaker
Ensure that Primary Configuration is selected to deploy a IBM Storage Scale Server active instance that can be used by IBM Storage Scale Clients.
Configure the following attributes:- Number of IBM Storage Scale Node(s)
- Specifies the number of IBM Storage Scale Servers to deploy.
- Cluster Name
- A name that defines this IBM Storage Scale cluster. Do not specify the text string IBM Storage Scale in the cluster name.
- File System name for selected shared volumes
- The file system name for the shared volumes (disks) that are used by the IBM Storage Scale shared file system. Do not specify the text string IBM Storage Scale string in the file system name.
- Block size
- Required. Use this property to set the block size for the specified file system. The default
value is set to 256 K. This value cannot be changed after the file system is created. If you cannot
test your application performance with various file system
blocksize
settings, use the default of 256 K. For more information on the value to choose for your application, see: - leaseRecoveryWait (seconds)
- Use this property to change the value of the
leaseRecoveryWait
for the IBM Storage Scale server cluster. The minimum allowed value is 10 seconds and the default value is 35 seconds. For a one node deployment, theleaseRecoveryWait
is automatically set to 10 seconds; otherwise, the value is set to 35 seconds. Update this property with caution since lower values might result in data corruption when nodes in the cluster are down. Contact the IBM Spectrum Scale service team and read the IBM Storage Scale documentation to understand the implications before you update this value.You can change this property after deployment, by using the IBM Storage Scale server instance Change Configuration Administrative operation. Run the Change Configuration operation and set
leaseRecoveryWait=<new value>
. For more information, see Change Configuration operation.For more information about this property, see Tuning Parameters.
- failureDetectionTime (seconds)
- Use this property to change the value of the
failureDetectionTime
for the IBM Storage Scale server cluster. The minimum allowed value is 10 seconds and the default value is 35 seconds. Update this value with caution. Contact the IBM Storage Scale service team and read the IBM Storage Scale documentation to understand the implications before you update this value.You can change this property after deployment, by using the IBM Storage Scale server instance Change Configuration Administrative operation. The IBM Storage Scale server nodes must be down when this operation is invoked, so first run a Stop Nodes administrative operation. Next, run the Change Configuration operation and setfailureDetectionTime=<new value>
, and then run Start Nodes to restart the nodes. For more information about these operations, see:For more information about this property, see Tuning Parameters.
- Snapshot Interval (minutes)
- The IBM Storage Scale file system produces a snapshot every 60 minutes (the default interval). You can change this value before or after deployment, by using the instance operations available under Manage > Operations > GPFS_Manager > Create and Manage Snaphots. You can use these snapshot operations to enable or disable snapshots, update snapshot time, delete a snapshot, or produce a one time snapshot. For more information, see the Related links.
- Active Configuration (Mirror and Tie) section
- Use this optional section to deploy a IBM Storage Scale Server Primary
configuration that has a Mirror configuration and a Tiebreaker configuration
attached to it. You must already have a Mirror configuration and a
Tiebreaker configuration deployed on a Cloud Pak System Software for Power instance.
You need to obtain the IP addresses for the IBM Storage Scale Manager virtual
machine from both the Mirror and Tiebreaker configurations, and enter
them in the fields provided in this section.
You can also attach the Mirror and Tiebreaker configurations after this Primary configuration is deployed, by using the Add New Member operation available under Manage > Operations > GPFS_Manager > Manage Cluster. Refer to the Related links for more information on these types of configurations.
- Passive Configuration section
- This section is not supported in Cloud Pak System Software for Power.
- With the IBM Storage Scale Server component
selected in the canvas, configure the following settings in the properties
pane:
- Click Save.
Results
Your new virtual application pattern is saved. You can find it in the list of available virtual application patterns by selecting Patterns > Virtual Applications.
What to do next
After deployment you can perform additional operations to manage the IBM Storage Scale configuration as needed.