Эффективная ИТ-инфраструктура удаленных офисов и филиалов

Deliver a Hyperconverged Remote Office-Branch Office Deployment with Hitachi Unified Compute Platform HC

Tech Note

See how Hitachi Unified Compute Platform HC V240 (UCP HC) is one of the best options to deploy a hyperconverged infrastructure to remote office-branch office (ROBO) sites. This tech note delivers the following information:

• Validation of Hitachi Data Ingestor (HDI) deployment on Unified Compute Platform HC in a ROBO site
• Automated backup of CIFS shares for Hitachi Data Ingestor to Hitachi Content Platform (HCP) in a corporate data center
• With Unified Compute Platform HC, Hitachi Data Ingestor, and Hitachi Content Platform, provide a cost efficient, small foot print, and low maintenance ROBO site solution with user data protection

Hitachi Unified Compute Platform HC delivers compute, storage, virtualization, management and data protection in an agile, scalable, and resilient hyperconverged infrastructure. The Unified Compute Platform HC hybrid and all-flash solutions deliver a reliable platform for business applications, databases, analytics, virtual desktop infrastructure, development, operations, and remote and branch office (ROBO) deployments.

Hitachi Data Ingestor is an elastic and backup-free cloud on-ramp and filer that enables an advanced edge-to-core storage solution. Data Ingestor with Hitachi Content Platform (HCP) creates an integrated offering that provides distributed consumers of IT, such as ROBO or cloud storage users, with a seamlessly scalable, backup-free storage solution. Deployed with a minimal-footprint or as a virtual appliance, Data Ingestor sends data from the edge to a core infrastructure, employing advanced storage and data management capabilities.

This solution also uses Hitachi Unified Compute Platform 2000 For VMware vSphere:

This tech note focuses on Hitachi Data Ingestor deployed on Unified Compute Platform HC in remote offices with limited IT resources or staff on-site. The following are some of the main benefits of this solution:

• Scale the implementation, starting with small increments of cost-effective capacity and grow as needed.
• Deploy, configure, and manage the solution with minimal IT experience.
• Save significant operational overhead with a smaller data-center footprint with reduced power and cooling expenses.
• Reduce the cost and complexity of management, backup, archiving, and compliance at the ROBO site with Hitachi Data Ingestor.

Figure 1 shows a simple overview of this solution.

• Deploy the virtual machines for Hitachi Content Platform on Hitachi Unified Compute Platform 2000 in the corporate data center.
• Deploy the virtual machines for Hitachi Data Ingestor on Hitachi Unified Compute Platform HC in the ROBO site/ This provides the CIFS share for Microsoft® Windows® desktop users.
• Automatically backup the user data in the CIFS share for Data Ingestor to Content Platform.

Table 1 lists the test cases used to validate this solution.

TABLE 1. TEST CASES AND RESULTS SUMMARY

This document does not cover the following:

• Sizing information
• Best practice
• Implementation details

For implementation details, please contact your Hitachi Vantara representative.

Note — Testing of this configuration was done in a lab environment. Many factors affect production environments beyond prediction or duplication in a lab environment. Follow the recommended practice of conducting proof-of-concept testing for acceptable results in a non-production, isolated, test environment that matches your production environment before your production implementation of this solution.

Tested Components

Table 2 shows the main hardware components used to conduct the test.
Table 3 lists the major software components used.
TABLE 2. HARDWARE COMPONENTS

Table 3 lists the major software components used.
TABLE 3. SOFTWARE COMPONENTS

Test Infrastructure and Configuration

Figure 2 shows the high-level infrastructure and main virtual machines deployed on a remote office and a data center.
Figure 2

The configuration of each site is as follows.

• • • Corporate Data Center
      • • Deploy four Hitachi Content Platform virtual machines with a standard configuration on Hitachi Unified Compute Platform 2000.
        • Create a Hitachi Content Platform tenant for the department. As an example, this tech note uses a tenant called Finance for the finance department.
          • Allocate a 900 GB quota.
  • Remote Office-Branch Office
  • From one Hitachi Unified Compute Platform HC, configure a four-node virtual SAN cluster.
  • Deploy one Hitachi Data Ingestor virtual machine on the Unified Compute Platform HC virtual SAN datastore. Connect this to the tenant called Finance on Hitachi Content Platform.
    Create a file system with the following settings to provide a CIFS shared folder called finance-shared for Microsoft Windows desktop users:
    • Namespace quota — 500 GB
    • Cache capacity — 100 GB
    • Enable file version restore and keep hourly versions for 8 hours. Keep small number of versioning for short period of time for testing purpose.
    • Migration/backup scheduling: every 30 minutes.

• Deploy multiple Microsoft Windows 10 desktop virtual machines on the Unified Compute Platform HC virtual SAN datastore, accessing the Hitachi Data Ingestor CIFS shared folder.

Table 4 lists the basic configurations of the virtual machines.
TABLE 4. VIRTUAL MACHINE CONFIGURATION

Generate New Data for Back Up for Testing

The user’s new data on the Hitachi Data Ingestor CIFS shared folder is automatically backed up to the corresponding Hitachi Content Platform tenant.
To observe and test this behavior, new data was generated constantly to the Hitachi Data Ingestor CIFS shared folder.
The following was used to generate new user data for this test:

• Use a Microsoft Windows virtual machine to access the finance-shared CIFS folder hosted by Hitachi Data Ingestor.
  • Over 100 files were created in various file sizes from 1 KB to 50 MB with a total of 1.3 GB used.
  • Among these files, 50 files were written regularly by a load generator.
  • Random data that is 64 KB in size is written to these files to generate over 3 GB of new data per hour.

• Back up the new data to the Hitachi Content Platform tenant every 30 minutes.

Test Result

This is a discussion of the test results in detail.

Deploy the Virtual Machines

This test involved the following:

• Deploy a Hitachi Data Ingestor virtual machine on Hitachi Unified Compute Platform HC in a remote site.
• Deploy Hitachi Content Platform virtual machines on Hitachi Unified Compute Platform 2000 on a data center site.

These are the results.

• The Hitachi Data Ingestor virtual machine was successfully deployed on the Hitachi Unified Compute Platform HC virtual SAN datastore.
• Four Hitachi Content Platform virtual machines nodes were successfully deployed on Hitachi Unified Compute Platform 2000.

Perform Automated Backup of a CIFS Share Hosted by Hitachi Data Ingestor

For this test, new data on the Hitachi Data Ingestor CIFS share was automatically backed up to the Hitachi Content Platform tenant every 30 minutes. To observe this, the performance monitor from the VMware vSphere web client was used.

Figure 3 shows the data transmit rate of the Hitachi Data Ingestor virtual machine. The data was transmitted from Hitachi Data Ingestor every 30 minutes.

Figure 3

Restore User Data from Hitachi Content Platform

If a file is deleted accidentally, or it is corrupted, users can restore these files from the .history folder. Under the history folder, there is a list of folders with timestamps, as shown in Figure 4. The user can restore the desired version of the file from these folders.

Figure 4

Self-sustained Hitachi Content Platform Storage Usage

During the Hitachi Data Ingestor file system creation, when enabling the file version restore option with keep hourly versions for 8 hours, the Content Platform tenant policy automatically enables versioning with Purge versions older than 2 days. This means any new writes and delta changes are kept in the Content Platform tenant quota for 2 days, and, after that, they are automatically deleted.

For this test, the load generator was kept running for 7 days with over 3 GB per hour of new data backed up to Content Platform. After 2 days, the storage usage of the Content Platform tenant was maintained between 130 GB and 200 GB. This indicates that Hitachi Content Platform storage can be self-sustained if the user data change rate does not increase.