Topology

Topology is a detail view included in the VMFS type Virtual Machines screen. The Topology view displays information collected for a Virtual Machine in a Service Processor/HPE Primera storage system along with the storage data from HPE SSMC.

Mandatory prerequisites for HPE SSMC to display VM-related Topology information

  • Configure VMware vCenter to the HPE Primera storage system or the HPE 3PAR Service Processor.

  • The VMWare vCenter must be managing virtual machines running on datastores that are created on relevant HPE 3PAR or Primera virtual volumes.

  • Configure Service Processor in HPE SSMC (for HPE 3PAR 3.3.1 only).

  • Ensure proper connectivity and time synchronization among HPE SSMC, Service Processor/HPE Primera storage system, and the VMware vCenter.

  • Ensure that the arrays are managed by SP 5.0.5 version and above.

  • Enable Topology in the Application Settings. Also, enable Analytics if applicable.

The Topology view displays information that is collected for the following components:

  • VM: VMFS Virtual machine that you select. The VM box displays the CPU usage and the amount of guest physical memory that is actively used.

  • VMDK: Virtual Machines Disk (VMDK) related to the Virtual Machine. The VMDK box displays the Read and Write throughput and Read and Write latency.

  • Application Type: In the Virtual Volume component, Application Type appears with the Appset_Name for related volumes that are associated with the App Volume Set. The same behavior applies to the VM component. Hover over the Application icon when there are multiple App Volume Sets.

  • Datastore: Datastore on which a VMDK resides. The Datastore box displays the following:

    • Device Latency: The average time required to Read or Write from or to the physical device.

    • Kernel Latency: The average time consumed by VMkernel to process each SCSI read and write command.

    • Queue Latency: The average time consumed in the VMkernel queue for each SCSI read and write command.

    • Throughput: The average amount of data that is read or written from or to the disk per second.

  • Host: ESXi host on which the datastore resides and the corresponding hosts in the storage system. The outer box displays the physical ESXi host. The inner box represents the host names by which the related ports are added in the storage system. A host if found to be a relative outlier by latency based on the Advanced analytics feature, is demarcated with a blue color. Port icon tooltip displays the FC port WWN/iSCSI IQN information. This correlation may not appear appropriately if there is a mismatch in port details across physical host & the ports added as in storage system.

    The ESXi host information displays the following:

    • CPU Usage: Represents the percentage of CPU used by all VMs and Host OS out of the total CPU available to host (number of CPU cores times the per core processor speed).

    • Active Memory: Represents the percentage of physical active memory for all powered-on VMs and vSphere services like CoS and vpxa running on host out of total memory available for host.

    Storage host (ports) displays the following:

    • Health status Tooltip on the health status displays the health status description.

    • Read and Write throughput, latency based on all its exported volumes.

    • Bulb icon appears next to Latency when the host is identified to be a relative outlier by its performance as per advanced analytics feature (If this is enabled in SSMC and for all flash storage systems only). A click on the icon opens a correlation window to view the performance trend.

  • Virtual Volume: Virtual volumes associated with the datastore. If applicable, displays the related App volume set name and type. The component displays the throughput and latency of:

    • Existing virtual volumes (VV): Based on the storage components such as physical disks.

    • Exported virtual volumes: Based on the data paths through which the volume is exported.

    The virtual volume displays the following information:

    • Health status.

    • Capacity used. The percentage of user space used capacity over the virtual size of the virtual volume.

    • Read and Write throughput and latency (VV) based on virtual volume statistics. For example, data from the storage back-end components like physical disks. For HPE Storage OS 3.2.2 MU4 and MU6 versions, the values for Throughput (VV) and Latency (VV) are not displayed in the VM Topology.

    • Read and Write throughput and latency (exported) based on the VLUN statistics. For example, data from paths on which volume is exported.

    • Bulb icon for latency (exported) appears if the virtual volume is identified to have unexplained high latency as per the advanced analytics feature (If enabled in SSMC and for all flash storage systems only). Click the bulb icon to open the correlation window to view the performance trend.

    • Volume, if found to have unexplained high latency as detected by the Advanced Analytics feature of SSMC is marked as blue and the same reflects in the timeslider.

  • System: Storage system from which the virtual volume is exported. The system information displays the following:

    • Health status. Tooltip on the health icon displays the health summary.

    • Firmware version.

    • Model.

    • Capacity allocated %

    • CPU usage %

    • Latest Saturation % as derived by the advanced analytics feature (if this is enabled in SSMC and for all flash storage systems only).

  • Timeslider:

    A time slider is included at the top of the view. The time slider helps to navigate back in time and see the configuration hierarchy, performance, and status aspects of the virtualization to storage topology objects. The time slider corresponds to the time for which the data is available for a maximum of seven days. You can view the topology information from the current time to the past seven days. The time stamp is based on the time zone of the storage system. To properly correlate the data, readings must be taken for the SSMC appliance, Storage System, Service Processor, vCenter, ESXi host, and Virtual Machine in the same time zone. If clocks across respective entities are synced with Network Time Protocol (NTP) server, then correlation can be done easily and the data will appear appropriately.

    Sliding over a time period helps you to figure out the:

    • Configuration changes over the selected time period.

    • Performance of a component during the selected time period and at a particular time.

    If an issue occurs at the selected point in time or time period, the color coding on the time slider enables you to understand the issue. You can troubleshoot the issue with the help of available Topology insights.
    NOTE:
    If the virtualization data is not available at a particular point, the latest time stamp may not be available in the time slider. The reasons for the unavailability of data are:

    • Unable to reach Service Processor/HPE Primera storage system from HPE SSMC, or vCenter from Service Processor/HPE Primera storage system.

    • Topology-related data collection disabled in HPE SSMC or in the Service Processor/HPE Primera storage system.

    Check the vCenter status in Systems > Settings pane and in Service Processor/HPE Primera storage system. Details related to added ESXi host may take up to one day to appear. After the data load, an auto refresh does not occur by default. Refresh the page to see data with latest time stamp.

    Color coding on the time slider

    The time slider has the following two divisions.

    • Top division is blue to depict the data availability from vCenter. Otherwise, it is white.

    • Bottom division corresponds to the storage side of the data.

    The different colors on the time slider points to the criticality of one of the storage components. The colors on the time slider and the bar indicate:

    • Red: Any storage component health is critical or the saturation level of storage system is in critical zone.

    • Orange: Any storage component health is in warning state, degraded, or saturation level of storage system is in warning zone.

    • Royal Blue: Host is identified to be relative outlier or volume is identified to have unexpected high latency. This is applicable only if the advanced analytics feature is enabled in SSMC and for all Flash storage systems only.

    • Green: All the resources are in a normal state.

The VM, VMDK, Datastore, and Host (ESXi) data that appears in the view is collected from the vCenter by Service Processor/HPE Primera storage system.

Connection hierarchy

The connection of the topology components is as follows:

  • VM to VMDK.

  • VMDK to Datastore.

  • Datastore to the ESXi Host.

  • Datastore to virtual volumes.

  • Storage system host to virtual volumes.

  • Virtual volumes to Storage system.