Workload Data Importer

The Workload Data Importer is designed to import and analyze workload data and then create a temporal workload model out of the imported workload data. The workload data can be exported from storage vendor system's IO stats or storage monitoring tools.

The data you import into WorkloadWisdom is in .csv format. The data is analyzed to understand the workload characteristics in terms of IOPs, throughput, read/write ratio, command distribution, and so forth.

Different storage systems collect different stats at different granularity, and the data that are collected are stored in different structures. WorkloadWisdom builds a dedicated analyzer and parser, called Analysis Policy, for each unique source of workload data. The Supported Analysis Policies section provides a full list of workload data sources that are supported.Supported Analysis Policies

For better manageability over time, imported workload data files are associated with Production Storage Environments. A Production Storage Environment (PSE) is a logical repository for one or more imported workload data files, and you should name it such that at a later date, you can easily find workload data for a specific storage environment that you care about. You can have, and are encouraged to have, multiple PSEs defined in WorkloadWisdom.

For example, if you are an IT Storage Engineer for your company's IT infrastructure, then you may have the following PSEs defined:

  • Tier 1 Applications

  • US West data center 2

  • NAS Infrastructure

  • Vendor X arrays

And if you are a QA engineer for your company's storage product team, then you may have the following PSEs defined:

  • Customer X Mission Critical data

  • Customer samples for regression resting

  • Support Escalation

When you import a workload data file, you choose the PSE that you want to associate the file with. At a later time, if you want to go back to some past workload data from US West data center 2 then it is easier to find it by browsing the PSEs you have created.

To get started, the following workflow is recommended, once you have the workload data file ready for import into WorkloadWisdom.

  1. Create Production Storage Environment

  2. Use Workload Data Importer to import the workload data file

  3. Wait for the analysis to complete, and review the analyzed workload data

  4. Create a workload model out of the analyzed workload data

  5. Run tests using the workload model

Notice

While the process of importing a supported workload data file into WorkloadWisdom is very simple and straightforward, much of the time may be spent collecting and extracting that data from a storage array in the first place. If this is your first time doing something like this, it is strongly recommended that you contact Virtana for help. Our Support team can help you with the general workflow and basics of extracting workload data, but it is beyond the scope of our standard product Support. For more hands-on and specific assistance, our Professional Services team can help train you and / or extract the workload data for you, so that you can become more familiar with the process of collecting workload data from your storage arrays.