Skip to content

setup.md

Latest commit

 

History

History
161 lines (121 loc) · 9.73 KB

setup.md

File metadata and controls

161 lines (121 loc) · 9.73 KB

OMS Workspace

An OMS (Operations Management Suite) workspace is needed as the platform aggregating data from all Azure Stacks.

To create an OMS workspace:

PRE-REQUISITES

  • Running instance of Azure Stack (1709 or higher)
  • Subscription to Azure (OMS Log Analytics)
  • Internet connection from Azure Stack OMS VM to Azure

Note: Windows Live ID credentials for Azure is not supported by this tool. So please make sure you deploy OMS using an account that is not a windows live ID.

SETTING UP THE LOG ANALYTICS RESOURCE

  1. Log into the Azure Portal (portal.azure.com)
  2. Go to Log Analytics via the search box
  3. Create a new log analytics resource by clicking the +Add button
  4. Provide the appropriate information.
  5. After the deployment succeeds, note down the resource group, Workspace Name, and Subscription ID. They will be needed when deploying the ARM template later.

Deploy Data Collection scheduled tasks to a data collection VM

To see aggregated usage from all of your Azure Stacks, you need to deploy a data collection VM on each of your Azure Stacks.

Prerequisites

  • A running Azure Stack that is registered to an Azure subscription that has downloaded Windows Server 2016 - eval image from Marketplace.
  • An OMS workspace.

Steps to Deploy

Step 1 - Install a VM in the Default Provider Subscription

  1. Login to admin portal
  2. Create Windows Server 2016 VM
  3. Name VM: UploadToOMSVM
  4. Choose A1 Standard
  5. Wait for VM deployment to complete

Step 2 - Prep the VM

  1. Download 2 files from the root of this github project to the c:\ of the VM you just deployed (InvokeMasterScript.ps1 and MasterScript.ps1)

Step 3 - Get required variables

The following are required to setup the environment. You should gather these variables before proceeding to the next step.

DeploymentGUID = “<e.g. 41da4fdd-0e5f-4ecb-85d2-52cb85cd1fca>”

  1. Access the privileged endpoint
  2. Run Get-AzureStackStampInformation
  3. Find and copy the deploymentguid from the output

azureStackAdminUsername ="<e.g. Serviceadmin@myazurestackinstance.onmicrosoft.com>"

  1. Update with the Azure Stack Service Admin account email

azureStackAdminPassword = "<e.g. MyAzureStackPassword206!>"

  1. Update with the Azure Stack Service Admin account password

CloudName ="<e.g. Orlando MTC>"

  1. Update location with the name of your Cloud, this is how most data will pivot in the views

Region = "<e.g. Orlando>"

  1. Update with the region name used when deploying Azure Stack

Fqdn = "<e.g. azurestack.corp.microsoft.com>"

  1. Update with the FQDN name used when deploying Azure Stack

azureSubscription = "Input your Azure Subscription GUID"

  1. Update with the Subscription ID used when deploying OMS

azureUsername ="<e.g. LoAnalyticsContributor@myazureinstance.onmicrosoft.com>"

  1. Update with the Azure account email with log analytics contirbutor role access to this log analytics workspace

azureStackAdminPassword = "<e.g. MyAzurePassword206!>"

  1. Update with the Azure account password for the Log Analytics contribtor role

OMSWorkpsaceName = "Name of your log analytics workspace"

  1. Update with the OMS Workspace Name OMSResourceGroup = ""
  2. Update with the OMS Workspace Resource Group name

Step 4 – Update variables

  1. Open an elevated PowerShell ISE session
  2. Open the file C:\InvokeMasterScript.ps1
  3. Update the variables using the data gathered in Step 3

Step 5 – Execute the script & update the scheduled task

  1. Run the InvokeMasterScript.ps1 now that the variables have been updated.
  2. Once the script completes, open Task Scheduler and update the Run As account to the Admin UserName and Password of the UploadToOMSVM VM.

The scripts sets up 2 scheduled tasks:

  1. Upload of 1-day worth of usage data provided from the Provider Usage API at 9am every day.
  2. Upload of operational data every 13 minutes.

The data are uploaded to the OMS workspace you specified in the ARM template.

Note: For usage data, the script is setup to query and upload usage data reported from the day before yesterday each time the scheduled task runs. Note that no usage data will be uploaded if there are no tenant usage during the timeframe specified.

Troubleshooting

  1. To verify that data is getting uploaded to OMS environment:

    Log in to your OMS workspace (after 1 day for usage data and 15 minutes for operational data) and click on all collected data in log search. If you can see operational and usage data there, the pipeline is working.

  2. What happens if I cannot find any data in OMS?

    There are a couple places things could have gone wrong.

    • The scheduled task(s) did not get set up. To check, rdp into the machine you deployed using the ARM template and launch Task Scheduler from Windows Start menu to check if there is a scheduled task called UsageDataUpload. If there isn't, it may help to check the logs for the custom script extension following question 1 step 3 and look for any errors with registering a scheduled task.
    • Unable to access the usage API. To check for this, look in C:\AZSAdminOMSInt\UsageSummary.json on the deployed VM, if there are no records here, that means usagesummaryjson.ps1 failed at one point.
    • There are no usage data collected. On the deployed VM, if C:\AZSAdminOMSInt\UsageSummary.json shows an empty array, that most likely means there were no usage data available in the time frame specified (the day before yesterday).
    • If there are usage entries in C:\AZSAdminOMSInt\UsageSummary.json and the sceduled task exists, then the error likely occurred in C:\AZSAdminOMSInt\uploadToOMS.ps1, which uploads usage data to OMS from the json file.

OMS Solution

IMPORTING THE OMS DASHBOARD

  1. Navigate to the newly created log analytics resource. From the OMS workspace overview page via the Log analytics overview in the azure portal, click on OMS portal.
  2. Select view designer
  3. Click Import and select a view from the Solution Tiles folder.

Note: Errors may appear since data has not been inputted. Once imported, say save and go back to the main dashboard.

  1. Repeat step 2-3 for the remaining views within the Solution Tiles folder.

CHECKING THE UPDATED OMS DASHBOARD

  1. To access the OMS dashboard, go back to the resource in the log analytics section, and click on OMS workspace.
  2. From there, click OMS portal, which should open a new tab and display the views currently uploaded to OMS.

Power BI

Prerequisites

  1. PowerBI Desktop. To download, see here: https://powerbi.microsoft.com/en-us/downloads/
  2. Usage data already piped up to your OMS workspace.

Installation Steps

  1. Download the usage_dashboard_template.pbit file (PowerBI template) from here.
  2. Open the file with Power BI desktop. PowerBI will complain that it cannot connect to the data sources specified. Next we adapt the existing queries to connect to the data sources we need.
  3. In the Home ribbon under External Data, click Edit Queries.
  4. Ensure Usage is selected in the left pane.
  5. In the View ribbon, open Advanced Editor.
  6. For the line
let Source = Json.Document(Web.Contents("..."))

You would want to modify the link such that it is referencing your OMS workspace in azure. To get the correct link, in the OMS workspace you created that has Azure Stack usage and operational data, go into log analytics and click Analytics tile in the top ribbon. In the new page, open a new tab by clicking + next to "Home Page" and enter the following query:

Usage_CL
| where Type_s == "Usage"
| project AdditionalInfo_s, DeploymentGuid_g , EndTime_t , Location_s , MeterID_g , MeterName_s , Quantity_d , ResourceName_s , ResourceType  , ResourceURI_s , StartTime_t , SubscriptionId , Tenant_s

and then select a time range in the top right corner. After that, click Export in the upper right corner, then click Power BI Query (M) to download a text file containing the a query for your workspace.

  1. Replace the content in the Advanced editor with the content of the file that you downloaded in the last step. Click Done.

  2. Click Edit Credentials in the banner that pops up and and navigate to the Organizational account tab to sign in using your Azure Credentials in the account that contains your OMS workspace. Click Save.

  3. Similarly, you can make capacity data available in PowerBI fetching it from OMS by repeating step 4 - 8, with a different query:

AzureStack_CL
| where Type_s == "Capacity"
| project DeploymentGuid_g , TimeGenerated, DiskAvail_d , DiskUsed_d , MemoryAvail_d , MemoryUsed_d , IPPoolAvail_d , IPPoolUsed_d

  1. There are two excel files containing supplementary information. You can find them here. Download them to your machine.

  2. In Power BI, in the Home ribbon, click on Data Source Settings. In the popup that follows, click on Change source... for each of the two excel files and change the source to point to the two excel files that you downloaded in the previous step.

  3. Click Close & Apply in the Home ribbon.

For specific documentation on the Power BI dashboard template, refer to the dashboard documentation.

Limitations

  1. As of 8/9/2017, there is 8mb size limit on the request that PowerBI uses to fetch data from OMS. This restriction is planned to be lifted in the future by the OMS Log Analytics team.
  2. If the restriction persists, one way to increase the number of days of usage data available (by 24x) is pull in daily aggregated usage data from OMS instead of hourly aggregated data. The obvious drawback for this method is that one will not be able to drilldown to the hourly level on the PowerBI dashboard.