Ensure that the system meets the minimum hardware configurations to install vRealize Network Insight.

Minimum Resource Requirements

  • vRealize Network Insight Platform OVA

    • 800 GB - HDD, Thin provisioned

    • Medium Brick Requirement

      • 8 cores - Reservation 4096 Mhz

      • 32 GB RAM - Reservation - 16GB

    • Large Brick Requirement

      • 12 cores - Reservation 6144 Mhz

      • 48 GB RAM - Reservation - 24GB

  • vRealize Network Insight Proxy OVA

    • 150 GB - HDD, Thin provisioned

    • Medium Brick Requirement

      • 4 cores - Reservation 2048 Mhz

      • 10 GB RAM - Reservation - 5GB

    • Large Brick Requirement

      • 6 cores - Reservation 3072 Mhz

      • 12 GB RAM - Reservation - 6GB

Software Requirements

  • Google Chrome or Mozilla Firefox Web browser

Privileges Required for Data Sources

  • Privileges required to configure and use IPFIX

    • vCenter Server Credentials with privileges:

      • Distributed Switch: Modify

      • dvPort group: Modify

    • The predefined roles in the vCenter server must have the following privileges assigned at root level that need to be propagated to the children roles:

      • System.Anonymous

      • System.Read

      • System.View

      • global.settings

  • Privileges required for NSX Manager Data Provider

    • NSX Manager Data Provider requires the Enterprise role.

    • If Central CLI is enabled, then the system admin credentials are required for NSX Manager Data Provider.

  • User privileges required on Cisco switches for metrics collection

    • vRealize Network Insight is capable of collecting metric data via SNMP as well as configuration via SSH from Cisco Switches. Cisco Switches UCS platform requires the use of both SSH and API for collection.

      Table 1.

      Type of data

      User Privileges

      Configuration Data

      Read-Only

      Metric Data

      SNMP read-only

      SNMPv2 read-only SNMP community

      SNMPv3 read-only

Optimum Performance Requirements

  • The maximum time skew has to be lesser than 30 seconds. The error message "One or more essential services are not running" appears.

  • The availability of the NTP service is critical to any system operations. Do not reboot the platform node if NTP service is not available.

  • The CollectorMain, IPfixProcessor, and Nginx services stop immediately if the proxy node reboots when the NTP server is unreachable. The error message " One or more essential services are not healthy" comes up.

    The workaround for this issue is to reboot the proxy node when the NTP server is reachable.

  • The vRealize Network Insight services do not recover automatically when the storage attached to the host is disconnected and rectified by addressing the storage connectivity issue.

  • The platform services crash and do not recover automatically when the existing compute resources are completely utilized by the other processes on the platform.

    The workaround for this issue is to reboot the platform node on which the services have failed to recover.

Brick Sizes

The hardware requirements of various brick sizes for a single platform and a single proxy VM are as follows:

Table 2.

Type

Brick Size

Capacity (Number of Managed VMs)

Flows (# of 4-Tuples)

Flow Records/IPFIX

vCPU Cores

RAM

Disk (Thin provisioned)

IOPS

Platform (With flows)

LARGE

6K

2M

12

48 GB

800 GB

250

Platform Without flows)

LARGE

10K

12

48 GB

800 GB

250

Platform (With flows)

MEDIUM

3K

1M

8

32 GB

800 GB

150

Platform ( Without flows)

MEDIUM

5K

8

32 GB

800 GB

150

Proxy (With flows)

LARGE

6K

100k/s

6

12 GB

150 GB

75

Proxy (Without flows)

LARGE

10K

6

12 GB

150 GB

75

Proxy (With flows)

MEDIUM

3K

50k/s

4

10 GB

150 GB

50

Proxy (Without flows)

MEDIUM

5K

4

10 GB

150 GB

50