ViPR SRM 3.6 – Manage Object Storage using ViPR SRM
Table of Contents
ViPR SRM supports the following reporting features for object storage:
- Consolidated Dashboard reports show available, free, and used object capacity across object storage arrays defined on Atmos, Centera, and ViPR software-defined storage.
- Consolidated Explore reports list all object arrays and show array details such as usage, free storage, object counts, and replication statistics.
- Additional Explore reports list existing object pools and details.
- The section includes performance reports for object arrays along with all other arrays. These reports show CPU utilization, memory utilization, or traffic problems at a glance for any array.
- The Report Library contains detailed health reports for the respective storage assets (Atmos, Centera, or ViPR software-defined storage).
- Consolidated Object Chargeback reports show object usage across all arrays for ViPR tenants and projects.
- In , you can edit rules for service levels related specifically to object arrays.
What's included in ViPR software-defined storage?
In the ViPR SRM enterprise-wide reports, the array type "Software-defined" refers to arrays that are defined using ViPR and consist of the following storage assets:
- Commodity hardware assets, such as approved HP server models with direct-attached disks, connected to ViPR nodes and managed as a virtual storage pool in ViPR
- ECS commodity hardware assets connected to an ECS appliance and managed as a virtual storage pool in ViPR
- Storage assets coordinated by the EMC Scale I/O solution and managed by ViPR
How is commodity storage represented?
Commodity storage connected to ViPR nodes and managed by ViPR software is represented as virtual pools. Using the ViPR user interface, an administrator assigns the virtual pool space into data stores. Data stores can be assigned to virtual arrays. Commodity storage connected to ECS also uses the data store concept.
What's included in a ViPR virtual data center?
A ViPR virtual data center can incorporate arrays from many sources in addition to the ViPR software-defined storage arrays. For example, a virtual data center might include a combination of file, block, and object arrays defined on devices such as Isilon, VNX, and Centera, as well as ViPR software-defined arrays backed by commodity hardware.
For details about ViPR virtual data centers and the arrays, pools, and storage systems that comprise a virtual data center, use the reports under. Back to Top
Assume that you are a system administrator with a request for 15TB of object storage from a new customer. You believe that your current collection of arrays on various devices is adequate. To provide solid confirmation, you use ViPR SRM to analyze object storage availability in your environment and research usage trends on existing object arrays.
- Log in to the Console and navigate to
In the summary capacity report, you see at a glance that your environment contains more than 750,000GB of configured and available object storage.
- To determine the arrays that contain the free object storage, navigate to
In this report, the Array Type column shows that your enterprise has object arrays defined on Centera and Atmos devices, in addition to a software-defined array managed by ViPR. Looking in the Free column, you see that all but the first array could satisfy your requirement for 15TB (approximately 15,000GB).
- To analyze and compare usage trends on the arrays, navigate to
This report includes a column for Utilization Percentage and another column that projects when the array will be 90% full based on current usage. You want to choose an array that has a favorable, long term projection.
The report shows that most arrays do not yet have configured pools and are not actively used.
Assume that you are administering an object storage program that charges some customers based on object counts and others by object usage. You want to use reports in ViPR SRM to track these numbers.
- Log in to the Console.
- To view the number of objects stored across the enterprise, by object array, navigate to
This report lists enterprise-wide object storage by array. The array types that you might see in this report include:
- object arrays defined in Centera clusters
- object arrays defined in Atmos clouds
- object arrays in ViPR software-defined storage using commodity hardware
- object arrays managed on ECS
The last column in this report shows Used Object Count for each object array.
- To view object counts and object access numbers for tenants and projects defined in ViPR, navigate to
This set of chargeback reports shows usage for objects on ViPR-managed commodity storage and on ECS-managed storage.
- Open one of the Capacity reports to see numbers of objects used by tenant or project.
- Open one of the Network Usage reports to see the number of times objects are accessed, by tenants and projects.
Assume that you are a System Administrator with operational responsibility for several important object storage customers. It is important for you to recognize and address potential performance backlogs before they present as real problems to your customers.
- Log in to the Console and navigate to .
- Explore the reports under Situations to Watch and Alerts for devices and components that might be experiencing problems. All device and array types are captured in these reports.
- To monitor Centera health:
- Navigate to .
- Notice the Offline column. You can drill into an offline cluster (array), if any, for more information.
- Navigate to an individual cluster name and then click Performance to display performance statistics.
Max Lookup Time, and review the trends.
For example, a consistent downward trend might need investigating.
- Navigate to , and review the Free and Faulted capacity columns.
- To monitor Atmos health:
- Navigate to Report Library and click EMC Atmos.
- Review the System Overview report, which shows defined cloud instances and their availability and disk usage percentages.
- Check disk health using the Top-N Reports.
Here is the Atmos report tree.
Trending and Forecasting Analysis and evaluate displayed values for metrics such as CPU utilization, memory, and total disk space.
Here is an example Memory Utilization report showing one node with high memory usage.
- To check the health of ViPR virtual data centers that might include software-defined object arrays:
- Navigate to
Two reports on that page distinguish between the ViPR cluster health and the supporting ViPR hosts:
- The Virtual Data Centers report shows the health of the ViPR cluster that represents the virtual data center.
- The ViPR Nodes report shows health for the ViPR servers.
- Navigate to
and explore the reports under that node.
Here is the ViPR report tree.
- To view alerts for all storage devices managed by ViPR, navigate to the global alerts area: .
- Navigate to .
The out of the box object service level names are the same as those used for LUN service level groups, but the criteria rules for the two sets of service levels are not related or connected.
The initial object service levels are based upon the following:
- The name of the ViPR virtual pool that the object bucket was provisioned from.
- The number of sites that the object is replicated across.
The object service level reporting is by tenant and project.
- Review the grading criteria for the out of the box object service levels and customize the rules if needed.
- For accounting and chargeback purposes, ensure that one of your object storage customers (the SRM Texas project) is consistently assigned to the Platinum service level.
- Log in to the Console, and click Administration in the banner.
- In the Administration navigation tree, click .
- In the Group Name list, click
Object Service Levels.
The Service Level definition page opens, showing the current members of the Platinum group.
- To view the service level criteria, click
View All Service Level Rules.
The initial criteria are similar to the following image.
In the rules statements:
- virtual pool name (vspname)
- Impacts the service level if it contains the service level name. For example, if the bucket was provisioned out of a virtual pool with the name SRM Texas Platinum Object vPool, the Platinum service level is assigned to that space.
- Is essentially the number of virtual data centers that are linked together for object replication. A value of 1 represents an isolated virtual data center with only local object protection. A value of 2 represents two virtual data centers that replicate objects across the two sites. A value of 3 means three sites, and so on.
- The higher the numsites value, the better the protection. Therefore, the better service levels have higher numsites values.
- To add a rule that assigns the SRM Texas project to the Platinum service level regardless of number of sites:
- Close the Service Level Rules page by clicking OK in the lower right corner of the page.
- On the main page, click Platinum, and then click Edit.
- On the Edit Service Level page, click
Add new rules set.
- Scroll to the bottom of the page to see the wizard that can help you create the new rule.
- In Select a type, choose Advanced.
- Move the cursor to the next field, and type the database field name (or a partial value) that you want to use in the rule. In this case, type
pr (for project name).
The wizard lists valid field names that begin with pr.
- Scan the resulting list, and select prjname.
- Select an operator. In this case, select Is for exact equal.
- Type the value that you want the rule to match. In this case, type SRM Texas.
The main page redisplays showing the current members of the Platinum group.
- Compare the current Platinum members to the original Platinum members in Step 3.