Configuring and Using the Profiler Application

Adding a Server to the Profiler

Adding vCenter to the ESXi Profiler

At the first login following installation of Hx Workload Profiler, you are redirected to the landing page where you can find the + Add Workload option on the top right corner of the page. You can select vCenter, then you can add multiple vCenters.

To calculate the metrics for a host, the Profiler captures the metrics for all the VMs on the Host. You then need to select the VM to be profiled. By default none of the VM’s are selected.

The workflow includes:

  • Node details tab: Records your vCenter details and then connects to the vCenter. If the connection succeeds, the Next button displays enabling you to select the hosts for profiling.

  • Select Hosts tab: Provides the vCenter details in hierarchy of vCenter, Datacenters, Clusters, and Hosts, with the option to select hosts to Profile. Default: All hosts in vCenter are selected.

  • Profiling Setting tab, which provides:

    • Profile Name: for specifying the name of the profile which must be unique.

    • Profile period: for specifying the period of profiling. Default value is 7days and the minimum value is 1day.

    • Polling Interval: for specifies the frequency for which polling should occur.

    • Quick Profile: for a quick estimate

    • Detailed Profile: for detailed end-to-end profile results.

Procedure

Step 1

Log in to the Profiler. See Logging in to the ESXi Profiler.

Step 2

Follow the steps provided to enter values for the following options:

Item

Description

vCenter Name

Name of the vCenter you are adding.

User Name

Name of the user as part of the login credentials for the vCenter.

Note 

HxProfiler uses VMware SDK to collect different performance metrics. This requires connecting to the VMware vSphere server using an user account that has either administrator privileges or with Server Manager role, which has the access privileges to get the performance data.

Password

The password set as part of the login credentials for the vCenter.

Polling Interval

The interval at which you want polling to occur. The default is 20 seconds. You can change the interval to between 20-120 seconds, based on the number of hosts and VMs being polled in that vCenter.

Step 3

Click Start Profiling. The Profiler service starts automatically.

When the connection status indicates successful connection, click Next.
Step 4

On the Poll filter tab, review the hierarchy values.

Adding Windows Bare Metal to the Profiler

At the first login following installation of Hx Workload Profiler, you are redirected to + Add Workflow. You can select Windows Bare Metal to add Windows Bare Metal to the workload.

SUMMARY STEPS

  1. Log in to the Profiler.
  2. On the Node details tab, enter values for the following options:
  3. If you select the Host Name/IP address option, click Save. If the given credentials / IP is valid, then workload will be added successfully. In case of any error, recheck and update the provided details, then click Save to add the workload successfully.
  4. If you select the IP Address Range, click Validate. Given IP range and credentials are validated, and if any of the VMs are not reachable, any error details will be updated in tabular format.
  5. If no errors are observed, click Save. If there are errors, please check the respective error VM details, rectify them manually and then click on Revalidate. Click Save to add the workload.

DETAILED STEPS

Step 1

Log in to the Profiler.

Step 2

On the Node details tab, enter values for the following options:

Item

Description

Windows Cluster Name

Name of the cluster you are trying to add
Host Name/IP Address

To enter a single host / different IPs

IP Address Range

To enter a range of VM IPs

User Name

Login username for windows baremetal

Password

Login password for windows baremetal

Polling Interval (seconds)

Default 20 seconds

Step 3

If you select the Host Name/IP address option, click Save. If the given credentials / IP is valid, then workload will be added successfully. In case of any error, recheck and update the provided details, then click Save to add the workload successfully.

Step 4

If you select the IP Address Range, click Validate. Given IP range and credentials are validated, and if any of the VMs are not reachable, any error details will be updated in tabular format.

Step 5

If no errors are observed, click Save. If there are errors, please check the respective error VM details, rectify them manually and then click on Revalidate. Click Save to add the workload.

Adding Linux Bare Metal to the Profiler

At the first login following installation of Hx Workload Profiler, you are redirected to + Add Workflow. You can select Linux Bare Metal to add Linux Bare Metal to the workload.

SUMMARY STEPS

  1. Log in to the Profiler.
  2. On the Node details tab, enter values for the following options:
  3. If you select the Host Name/IP address option, click Save. If the given credentials / IP is valid, then workload will be added successfully. In case of any error, recheck and update the provided details, then click Save to add the workload successfully.
  4. If you select IP Address Range, click Validate. Given IP range and credentials are validated, and if any of the VMs are not reachable, any error details will be updated in tabular format.
  5. If no errors are observed, click Save. If there are errors, please check the respective error VM details, rectify them manually and then click on Revalidate. Click Save to add the workload.

DETAILED STEPS

Step 1

Log in to the Profiler.

Step 2

On the Node details tab, enter values for the following options:

Item

Description

Linux Cluster Name

Name of the cluster you are trying to add
Host Name/IP Address

To enter a single host / different IPs

IP Address Range

To enter a range of VM IPs

User Name

Login username for Linux baremetal

Password

Login password for Linux baremetal

Polling Interval

Default 20 seconds

Step 3

If you select the Host Name/IP address option, click Save. If the given credentials / IP is valid, then workload will be added successfully. In case of any error, recheck and update the provided details, then click Save to add the workload successfully.

Step 4

If you select IP Address Range, click Validate. Given IP range and credentials are validated, and if any of the VMs are not reachable, any error details will be updated in tabular format.

Step 5

If no errors are observed, click Save. If there are errors, please check the respective error VM details, rectify them manually and then click on Revalidate. Click Save to add the workload.

Adding Hyper-V to the Profiler

At the first login following installation of the Hx Workload Profiler, you are redirected to the landing page where you can find the+Add Workload option on the top right corner of the page. Select HyperV to add multiple HyperV.

To calculate the metrics for a host, the Profiler captures the metrics for all the VMs on the Host. Select the VM to be profiled (by default none of the VMs are selected).

The workflow includes:

  • Node details tab: Records your Hyper-V details and then connects to the Hyper-V. If the connection succeeds, details regarding the profile named appear. You can also specify and save the profile period time for profiling by clicking on the start profile button to start profiling.

Procedure

Step 1

Log in to the Profiler. See Logging in to the Hyper-V Profiler.

Step 2

On the Node details tab, enter values for the following options:

Item

Description

Hyper-V Cluster Name

Name of the Hyper-V you are adding.

Host IP

IP address of the Hyper-V node. The Add button allows you to add multiple Hyper-V host IP addresses.

User Name

Name of the user as part of the login credentials for the Hyper-V.

Note 

The User Name must have the Admin role assigned to it.

Password

Password being set as part of the login credentials for the Hyper-V. The Add button allows you to add all the username and passwords for the IPs mentioned in the Host IP section.

Note 

If the username and password is same for all the Hyper-Vs or any 2 or more Hyper-Vs, then there is no need to mention it again.

Polling Interval (seconds)

The interval at which you want polling to occur. Default: 20 seconds. You can change the interval to between 20 and 120 seconds.

Step 3

Click Save.

Starting Data Profiling

Starting ESXi Data Profiling

Following the successful addition of a vCenter Server, the new vCenter displays on the Data Inventory (home) page. You configure the profiling attributes by providing the profile name and duration.

Table 1. Profiling Operation Options

Item

Description

Delete (trash can)

Deletes a previously added vCenter.

Edit (pen)

Edit vCenter properties to add or remove hosts for polling.

Stop ( || symbol)

Stops the profiling so you can resume it later.

Reset (refresh symbol)

Performs a reset operation, which creates a new profile and starts polling. When you trigger reset, the profiler stops the active/running profile and creates a new one. A prompt asks for confirmation.

View Collection

Opens the View Collection page so you can browse through the collected data as part of the profiling to review the HOST and VM level data.

Polling starts as soon as you enter the profiler name ,days and polling period.

Procedure

Step 1

In the dialog that displays, perform the following steps:

  1. Enter a name for the profile.

  2. Select a duration value from the Profiling Period down-drop list.

  3. Click Ok.

Following successful profile creation, the Profiler begins polling the selected hosts and all the VMs on those hosts. When the polling starts, the data collector runs as a background process. The Datacenter Inventory page displays information about the hosts and polling, showing number of hosts and the status of the polling.
Step 2

(Optional) To stop an in-progress profiling operation so you can resume it later, see Stopping the Profiler Service.

Step 3

(Optional) To stop an in-progress profiling operation, click Reset. You can then start a new one.

Step 4

(Optional) To browse through the collected data, see Viewing an ESXi Collection.

Step 5

(Optional) To download profiling data, see Downloading ESXi Profiling Results.

Starting Windows Bare Metal Data Profiling

Following the successful addition of a Windows Bare Metal, the newly added Windows Bare Metal displays on the Data Inventory (home) page. You can configure the profiling attributes by providing the profiling name and duration.

Table 2. Profiling Operation Options

Item

Description

Delete (trash can)

Deletes the selected Windows Bare Metal workload / node

Edit (pen)

Edit Windows Bare Metal to add / remove existing VM/hosts for profiling.

Stop ( || symbol)

Stops the profiling

Refresh

Performs a reset operation, which creates a new profile and starts the polling. When reset is clicked, profiler stops the active/running profile and creates a new one. A prompt asks for confirmation.

View Collection

Opens the View Collection page so you can browse through the collected data as part of the profiling to review the HOST and VM level data.

SUMMARY STEPS

  1. (Optional) To stop an in-progress profiling operation so you can resume it later, see Stopping the Profiler Service
  2. (Optional) To stop an in-progress profiling operation, click Reset. You can then start a new one.
  3. (Optional) To browse through the collected data, see Viewing an ESXi Collection.
  4. (Optional) To download profiling data, see Downloading ESXi Profiling Results.

DETAILED STEPS

Step 1

(Optional) To stop an in-progress profiling operation so you can resume it later, see Stopping the Profiler Service
Step 2

(Optional) To stop an in-progress profiling operation, click Reset. You can then start a new one.

Step 3

(Optional) To browse through the collected data, see Viewing an ESXi Collection.

Step 4

(Optional) To download profiling data, see Downloading ESXi Profiling Results.

Starting Linux Bare Metal Data Profiling

Following the successful addition of a Linux Bare Metal, the newly added Linux Bare Metal displays on the Data Inventory (home) page. You can configure the profiling attributes by providing the profiling name and duration.

Item

Description

Delete (trash can)

Deletes selected Linux Bare Metal workload / node

Edit (pen)

Edit Linux Bare Metal to add / remove existing VM/hosts for profiling.

Stop ( || symbol)

Stops the profiling

Refresh

Performs a reset operation, which creates a new profile and starts the polling. When reset is clicked, profiler stops the active/running profile and creates a new one. A prompt asks for confirmation.

View Collection

Opens the View Collection page so you can browse through the collected data as part of the profiling to review the HOST and VM level data.

SUMMARY STEPS

  1. (Optional) To stop an in-progress profiling operation so you can resume it later, see Stopping the Profiler Service
  2. (Optional) To stop an in-progress profiling operation, click Reset. You can then start a new one.
  3. (Optional) To browse through the collected data, see Viewing an ESXi Collection.
  4. (Optional) To download profiling data, see Downloading ESXi Profiling Results.

DETAILED STEPS

Step 1

(Optional) To stop an in-progress profiling operation so you can resume it later, see Stopping the Profiler Service
Step 2

(Optional) To stop an in-progress profiling operation, click Reset. You can then start a new one.

Step 3

(Optional) To browse through the collected data, see Viewing an ESXi Collection.

Step 4

(Optional) To download profiling data, see Downloading ESXi Profiling Results.

Starting Hyper-V Data Profiling

Following the successful addition of a Hyper-V Server, the new Hyper-V displays on the Data Inventory (home) page. You configure the profiling attributes by providing the profile name and duration.

Table 3. Profiling Operation Options

Item

Description

Delete (trash can)

Deletes a previously added Hyper-V.

Edit (pen)

Edit Hyper-V properties to add or remove hosts for polling.

Stop ( || symbol)

Stops the profiling so you can resume it later.

Reset (refresh symbol)

Performs a reset operation, which creates a new profile and starts polling. When you trigger reset, the Profiler stops the active/running profile and creates a new one. A prompt asks for confirmation.

View Collection

Opens the View Collection page so you can browse through the collected data as part of the profiling to review the HOST and VM level data.

Procedure

Step 1

In the dialog that displays, perform the following steps:

  1. Enter a name for the profile.

  2. Select a duration value from the Profiling Period down-drop list.

  3. Click Ok.

Following successful profile creation, the Profiler begins polling the selected hosts and all the VMs on those hosts. When the polling starts, the data collector runs as a background process. The Datacenter Inventory page displays information about the hosts and polling, showing the number of hosts and the status of the polling.
Step 2

(Optional) To stop an in-progress profiling operation so you can resume it later.

Step 3

(Optional) To stop an in-progress profiling operation, click Reset. You can then start a new one.

Step 4

(Optional) To browse through the collected data, see Viewing a Hyper-V Collection.

Step 5

(Optional) To download profiling data, see Downloading Hyper-V Profiling Results.

Downloading Profiling Results

Downloading ESXi Profiling Results

Procedure

Step 1

On the Datacenter Inventory page, locate and select the profile whose data you want to download.

Step 2

Use the Download option to select one of the following:

Option Description
All Data Collection

Downloads 30-day host summary, time series, and CVS for both the host and VM

Summary Report for HxSizer Upload

 Collects the historic 30-day information from vCenter and does not require any active polling. The download provides the output in CSV format and can be directly uploaded to the Compute and Capacity Workload of HxSizer.
Profiler Sample Data Downloads the sampled data for the selected profile in the following formats:

  • Summarized host data (CVS)

  • Summarized VM data (CVS)

  • Time series data of host (zipped CSV file)

  • Time series data of VM (zipped CSV file)
Profiler Summary PDF

PDF download

Step 3

You can still download the profiling results when viewing the compute, storage and network data of various hosts and VMs by clicking on the Download button on the top right corner of the UI.

Downloading Hyper-V Profiling Results

Procedure

Step 1

On the Datacenter Inventory page, locate and select the profile whose data you want to download.

Step 2

Right-click the Download icon at the upper right of the page to select to download one of the following:

Option Description
All Data Collection

Downloads time series and CVS for both the host and VM
Sizer Upload Summary Collects the host summary information from Hyper-V. The download provides the output in CSV format and can be directly uploaded to the Compute and Capacity Workload of HxSizer.
Profiler sampled data Downloads the sampled data for the selected profile in the following formats:

  • Summarized host data (CVS)

  • Summarized VM data (CVS)

  • Time series data of host (zipped CSV file)

  • Time series data of VM (zipped CSV file)
Profiler summary

PDF download

Step 3

You can still download the profiling results when viewing the compute, storage and network data of various hosts and VMs by clicking on the Download button on the top right of the UI.

Viewing Data Collections from Servers

Viewing ESXi Data Collections from vCenter Servers

The View Collection page has five tabs at the top left of the page, Overview, Host Compute Metrics, Host Storage Metrics, VM Compute Metrics, VM Storage Metrics. When clicked, they show summary data described in his section. The View Collection page also provides the summarized data for Compute Summary and Storage Summary of individual host and VMs. You can also fetch the data for specific period of profiling using the predefined filter present on the top right corner of the page with the minimum being 30minutes or the user can also use the time line to select the time period. The following sections describe the summary data shown through filter tool use and in that shown in each tab and view.

Table 4. Host View Filter Options

Item

Description

Aggregation

Filters to view the summarizations based on peak or average selections. Your selection determines the display of the table metrics and trends. The summarized values represent the following:

  • Peak: Peak value of all the metrics in the selected interval

  • Average: Averages of all the values in the selected interval

You can filter Host metrics using the search option by host name only. For VM metrics, you can filter either by the Host name to which the VM belongs or directly with the VM name, using the search option.

Host and VM Summarized Metrics

Based on your time period and parameters selections, the Profiler computes and populates the compute and storage metrics. Metrics display for the following values:

Host View Compute Table

  • Host_Name

  • Processor Type

  • Clock (GHz)

  • #Physical Cores

  • HT – ON/OFF

  • CPU Util (MHz)

  • CPU Util (%)

  • #VMs

  • vCPU of active VMs/pCore Ratio

  • RAM (GB)

  • RAM Util (GB)

  • Network Throughput- Rx (Mbps)

  • Network Throughput- Tx (Mbps)

Host View Storage Table

  • Provisioned Capacity (TB)

  • Used Storage Capacity (TB)

  • Read Throughput (MBps)

  • Write Throughput (MBps)

  • Read (%)

  • Write (%)

  • Read IOPS

  • Write IOPS

  • Read Block Size (KB)

  • Write Block Size (KB)

  • Seq (%)

  • Read Latency (ms)

  • Write Latency (ms)

VM View Compute Table

  • VM Name

  • Status

  • Host_Name

  • vCPUs

  • CPU Util (MHz)

  • CPU Util (%)

  • RAM (GB)

  • RAM Util (GB)

  • Network Throughput- Rx (Mbps)

  • Network Throughput- Tx (Mbps)

VM View Storage Table

  • VM Name

  • Host_Name

  • Disk Capacity (GB)

  • Disk Used (GB)

  • Read Throughput (MBps)

  • Write Throughput (MBps)

  • Read (%)

  • Write (%)

  • Read IOPS

  • Write IOPS

  • Read Block Size (KB)

  • Write Block Size (KB)

  • Seq (%)

  • Read Latency (ms)

  • Write Latency (ms)

Host and VM Trends

The View Collection page provides trend charts and an overview of various parameters at a host level and VM level for both compute and storage parameters. The overview provides information aggregate storage and compute matrix along with a 30-day sizing summary. You can view the trend charts by selecting the host or the VM from the table.

Metrics for the following values display in the trend charts:

Host View Compute Trends

  • CPU Utilization (MHz)

  • CPU Utilization (%)

  • CPU Overprovisioning Ratio (%)

  • RAM Utilization (%)

  • RAM Overprovisioning Ratio (%)

  • Receive Rate (Mbps)

  • Transmit Rate (Mbps)

VM View Compute Trends

  • Receive Rate (Mbps)

  • Transmit Rate (Mbps)

Host and VM View Storage Trends

  • Read Throughput (MBps)

  • Write Throughput (MBps)

  • Read Ratio

  • Write Ratio

  • Read IOPS

  • Write IOPS

  • Read Seq (%)

  • Write Seq (%)

  • Read Latency (ms)

  • Write Latency (ms)

  • Provisioned Capacity (TB)

  • Used storage Capacity (TB)

Histogram Charts

  • Read Block Size Histogram

  • Write Block Size Histogram


Note

You can plot the storage parameter charts for two comparable charts either from the Trend Line chart or from the Histogram chart.

Viewing an ESXi Collection

You can also use the filter and search tool in the Hosts and VM views to display only those hosts and VMs that you want to see.

You can also use the filter tool in the Hosts and VM views to display only those hosts and VMs you want to see.

Procedure

Step 1

In the Datacenter Inventory page, click View Collection to browse through the collected data.

Step 2

In the View Collection page, select between Host Compute Metrics, Host StorageMetrics, VM Compute Metrics or VM Storage Metrics.

Step 3

In the Select VMs tab, select or unselect the toggle switch for the VMs you want to view, then click Next. All VMs are selected by default.

Step 4

To view data for a different time period, select from the options in the top right corner with the default minimum value of 30 minutes. Use the time slider above the fixed time period selection to view data for a specific time range from the selected time period.

Viewing Data Collections from Hyper-V Servers

The View Collection page has five tabs at the top left of the page, Overview, Host Compute Metrics, Host Storage Metrics, VM Compute Metrics, VM Storage Metrics. When clicked, they show summary data described in his section. The View Collection page also provides the summarized data for Compute Summary and Storage Summary of individual hosts and VMs. You can also fetch the data for a specific period of profiling using the predefined filter present on the top right corner of the page with the minimum being 30minutes. You can also use the time line to select the time period. The following sections describe the summary data shown through filter tool use and in that shown in each tab and view.

Table 5. Host View Filter Drop-Down Options

Item

Description

Aggregation

Filters to view the summarizations based on peak or average selections. Your selection determines the display of the table metrics and trends.

The summarized values represent the following:

  • Peak: Peak value of all the metrics in the selected interval

  • Average: Averages of all the values in the selected interval

You can filter Host metrics using the search option by host name only. For VM metrics, you can filter either by the Host name to which the VM belongs or directly with the VM name, using the search option.

Host and VM Summarized Metrics

Based on your time period and parameters selections, the Profiler computes and populates the compute and storage metrics. Metrics display for the following values:

Host View Compute Table

  • Host Name

  • Processor Type

  • Clock (GHz)

  • #Physical Cores

  • Hyper Threading – ON/OFF

  • CPU Util in %

  • #VMs

  • vCPU of active VMs/pCore Ratio

  • RAM (GB)

  • RAM Util (GB)

  • Network Throughput- Rx (Mbps)

  • Network Throughput- Tx (Mbps)

Host View Storage Table

  • Host Name

  • Provisioned Capacity (TiB)

  • Used Storage Capacity (TiB)

  • Read Throughput (MBps)

  • Write Throughput (MBps)

  • Read (%)

  • Write (%)

  • Read IOPS

  • Write IOPS

  • Latency (ms)

VM View Compute Table

  • VM Name

  • Status

  • Host Name

  • vCPUs

  • CPU Util (MHz)

  • CPU Util (%)

  • RAM (GB)

  • RAM Util (GB)

  • Network Throughput- Rx (Mbps)

  • Network Throughput- Tx (Mbps)

VM View Storage Table

  • VM Name

  • Host Name

  • Disk Capacity (GB)

  • Disk Used (GB)

  • Read Throughput (MBps)

  • Write Throughput (MBps)

  • Read (%)

  • Write (%)

  • Read IOPS

  • Write IOPS

  • Latency (ms)

Host and VM Trends

The View Collection page provides trend charts and an overview of various parameters at a host level and VM level for both compute and storage parameters. The overview provides information related to aggregate storage and compute matrix. You can view the trend charts by selecting the host or the VM from the table.

Metrics for the following values display in the trend charts:

Host View Compute Trends

  • CPU Overprovisioning Ratio (%)

  • RAM Utilization (GB)

  • RAM Overprovisioning Ratio (%)

  • Receive Rate (Mbps)

  • Transmit Rate (Mbps)

VM View Compute Trends

  • RAM Utilization (GB)

  • Receive Rate (Mbps)

  • Transmit Rate (Mbps)

Host View Storage Trends

  • Read Throughput (MBps)

  • Write Throughput (MBps)

  • Read Ratio

  • Write Ratio

  • Read IOPS

  • Write IOPS

  • Read Seq (%)

  • Write Seq (%)

  • Read Latency (ms)

  • Write Latency (ms)

VM View Storage Trends

  • Read Throughput (MBps)

  • Write Throughput (MBps)

  • Read Ratio

  • Write Ratio

  • Read IOPS

  • Write IOPS


Note

You can plot the storage parameter charts for two comparable charts from the Trend Line chart.

Viewing a Hyper-V Collection

You can view a summary of compute and storage parameters for the hosts and VMs available in the Hyper-V through the View Collections page as part of the profiling operation.

You can also use the filter and search tool in the Hosts and VM views to display only those hosts and VMs you want to see.

Procedure

Step 1

In the Datacenter Inventory page, click View Collection to browse through the collected data.

Step 2

In the View Collection page, select between Host Compute Metrics, Host Storage Metrics, VM Compute Metrics or VM Storage Metrics.

Step 3

In the Select VMs tab, select or unselect the toggle switch for the VMs you want to view, then click Next. All VMs are selected by default.

Step 4

To view data for a different time period, select from the options in the top right corner with the default minimum value being 30 minutes. Use the time slider above the fixed time period selection to view data for a specific time range from the selected time period.

30-Days Sizing Summary Report

Upload 30-Days Sizing Summary Report

You can upload the 30-Days Sizing Summary Report to the HxSizer application from HxProfiler. When the report upload is successful, a reference number is generated. View the history of uploads along with the reference number from the HxProfiler application. In addition to identifying the report, the refence number may be used to claim the scenario from the HxSizer application.

Perform the following steps in the Cisco HX Profiler dashboard to upload the 30-Days Sizing Summary report:

Procedure

Step 1

Click the Upload icon on the HX Profiler ribbon and select Upload To HXSizer. The Upload Summary Report To HXSizer window appears.

Step 2

Click the Browse button to navigate to the local .csv file that needs to be uploaded to the HXSizer portal.

  1. Select the desired .csv file.

  2. (Optional) To modify the csv file name, type a new name in the Rename selected .csv file name field.

  3. Click Upload.

A reference number indicates a successful upload. Use the Copy button to copy the reference number and use it to claim the scenario in the HXSizer application.

Step 3

(Optional) View Upload History

  1. Click the Upload icon on the HX Profiler ribbon and select Upload History to view the list of 30-Days sizing summary .csv files that have been uploaded to the sizer application from profiler.