In my previous posts I whowed how to deploy, discover physical arrays, create virtual arrays and pools, and create and authentication provider. Next I will show how to deploy and configure object data services.
From the web GUI select settings, data service nodes
The data services nodes screen pops up. There are 3 basic steps. Define the node by IP, configure the node, and deploy services. Click on Configuration Properties
In my lab i am deploying a VM ovf for my dataservices node. It will use the Isilon file array I previously setup to provide storage for object. The IP i assign here will be the same IP I assign the ovf VM during deployment. Enter the IP address in the Extra Node IP field and click Save. The ViPR controller will reboot and bring you back to the data service configuration page
Click OK to reboot the controller
After the reboot completes click Download Configuration File and save it to the directory where you have downloaded the ViPR datservices ovf
In virtual center click the deploy ovf link. The wizard will pop up. Browse to the ViPR dataservices ovf and click Next
Accept the licensing agreement. Click Next
Choose and inventory location. Click Next
Choose a host cluster. Click Next
Choose a resource pool. Click Next
Choose a data store. Click Next
Accept the disk format. Click Next
Choose a network to use. Click Next
Note that you should be able to communicate to the controller from this network
Enter the IP information for the data service VM. Note that the IP address should be the same configured on the controller for the data services. Click Next
Review your settings. Select Power on after deployment. Click Finish
After the deployment finishes back on the controller gui, the next step is to click Data Service settings. However we first need to setup a data store and Virtual object pool to house the object data.
Got to Virtual assets, Data stores
Create a data store and associate it with a file Virtual array and file virtual pool. Click Save
A data store is now created that will be used for object data.
Go to Virtual Assets and Object Virtual pool
Create a Virtual Object pool and associate it with a file virtual array. Click Save
An Object virtual pool is created. Now we can go to the Data Service Settings and configure it.
Switch from the admin role to the user role
Click tenant settings and projects
Create a project. Click Save
We now have a project. Projects get associated with tenant namespaces and contain the ACL to control provisioning accesss. Tenant namespaces are a 1 to one relationship with a project
Go to tenant settings, data service seettings
Create a tenant namaespace and associate it with a Virtual Pool and project. Click Save
We have now setup a Object data store, associated it with a tenant and are ready to provision a bucket.
As the user, click the services and choose the HDFS and Object based Storage Services tile
Choose Create a new bucket
Give the bucket a name (note that if you wish to use s3 tools, conform to their standards for bucket names HERE) and associate it with a virtual pool and project. If you wish to enable HDFS on this bucket select the Enable File System Support check box. Click Order
The bucket is created. To access it we can use something like s3 browser
To access the bucket
From the emc community page https://community.emc.com/docs/DOC-27823 we can find the networking guide for ViPR Data services. For applications to access ViPR s3 buckets we must create DNS or local host records. First we need an “A” record for our data services VM that was deployed to host ViPR data services. For this blog we have the following
Next we have to create a URL and generate a access key
As the admin user click Settings, Object Base URL
Give the URL a name and a base. Choose use base URL with namespace. Click Save
Click root, Object Access Keys
A key is generated. Copy it.
Install the s3 browser. Open it and click Accounts, Manage Accounts
Enter an account name, the Access Key Id (root) and the Secret Access Key (generated on the ViPR controller in previous step). Click Advance
Click Use Amazon s3 compatiable storage. Click Close
ViPR s3 is accessed on the following ports
- 9020: S3 (HTTP)
- 9021: S3 (HTTPS)
Click Add new account
Click Save changes
The s3 browser should now connect to the bucket.