Monday, 28 January 2019

Intune Support for Azure Monitor - Public Preview

This is a great new feature of Microsoft Intune (Public Preview). Central logging has been a request for some time. Now we can send Audit and Operational Intune logs to Azure Monitor services. This allows us to:
  • Archive Intune logs to an Azure storage account to keep the data, or archive for a set time.
  • Stream Intune logs to an Azure event hub for analytics using popular Security Information and Event Management (SIEM) tools, such as Alien Vault, Splunk and QRadar.
  • Integrate Intune logs with your own custom log solutions by streaming them to an event hub.
  • Send Intune logs to Log Analytics to enable rich visualizations, monitoring, and alerting on the connected data.
In this example I'll archive the data to a storage account for analysis later.

What do we need?
  • Azure subscription
  • Intune
  • Global Admin or Intune Service Administrator account
  • Azure storage account with ListKeys permission (you'd need a different Azure service depending on where you want to route the logs)
So let's configure this logging.

In the Azure portal navigate to Intune

Select Diagnostics Settings and turn it on.

We can see three options:
  • Archive to a storage account: Saves the log data to an Azure storage account. Use this option if you want to save or archive the data. (this is the one we will use)
  • Stream to an event hub: Streams the logs to an Azure event hub. If you want analytics on your log data using SIEM tools, such as Splunk and QRadar, choose this option.
  • Send to Log Analytics: Sends the data to Azure log analytics. If you want to use visualizations, monitoring and alerting for your logs, choose this option

See the event hub configuration (if this was our preferred option).......

....or the log analytics (if this was our preferred option).

We'll just choose an Azure storage account (our option).

Enter a descriptive name for the diagnostic setting and configure the retention period in days. Save the setting.

The diagnostic setting has been created. You can edit this afterwards.


Remember that there will be a cost associated with this logging. See the Microsoft docs for more details.

Reviewing the logs

After an action the corresponding log will show up in the storage account between 5 and 15 minutes later and it's very easy to review the logs.

Navigate to Azure Storage accounts

and select the storage account you configured earlier.

Select Storage Explorer (also in Preview). You'll see a new Blob container - "Insights-log-auditlogs"

Drill down into the container and you'll find a .json file.

You can open the file in a browser to see the logs. See that some managed devices were deleted. I wonder who did that.

The Azure Storage Explorer tool is another option and is also handy for browsing this content. Download it from here

Install the tool and sign in to Azure.

Choose your subscription and select Apply.

How you can browse and open the logs.

Look who deleted the managed devices. I'll have to talk to him.

This is a brilliant new feature. I hope this blog post helps. Until next time.....

Thursday, 24 January 2019

SCCM PXE Responder and DHCP

One of the most eagerly awaited features of ConfigMgr Current Branch 1806 was the ability to enable PXE on Distribution Points without Windows Deployment Services (WDS). This was an enormous step forward as it can save customers money. I have customers with many PXE-enabled DPs throughout the world. That's a lot of server licensing. Previously a server OS was required simply for WDS. Now we can use a desktop operating system for the DP and enable it to use the lightweight PXE responder.

We were implementing this solution for a customer back in December and we had some trouble getting it to work. We were enabling the PXE responder on a server OS which may sound unusual but the customer was concerned about a possible reboot required for WDS and the logistics of organising this on 100 sites worldwide. It should still have worked - right?

The “Configmgr PXE responder servicewas running successfully but we still saw this error in the SMSPXE.log file.

PXE: bind() failed for DHCP, 00:15:5D:05:16:00,, 67. 0x80072740.        SCCMPXE            12/20/2018 5:28:37 AM         6448 (0x1930)

This actually made sense as this server was also configured as a DHCP server. Of course that was the problem. Both DHCP and the PXE requests use port 67. This is ok when using WDS as we can configure it not to listen on port 67. Unfortunately we didn't have the same option with the PXE Responder.

I reported this to the product group on the 21st December. Other MVPs chimed in (thanks @jarwidmark) and talked about the merits of making this work. I'm pleased to say that this fix was released yesterday 23rd Jan in Technical Preview 1901, only a month later. That's an amazing turnaround for a feature. Hats off to @TweetKerwin and @djammmer and this shows the power of the ConfigMgr community. 

Details of the fix can be found in the TP1901 docs

When you enable a PXE responder on a distribution point without Windows Deployment Service, it can now be on the same server as the DHCP service. Add the following settings to support this configuration:
  • Set the DWord value DoNotListenOnDhcpPort to 1 in the following registry key: HKLM\Software\Microsoft\SMS\DP
  • Set DHCP option 60 to PXEClient
  • Restart the SCCMPXE and DHCP services on the server.

Hopefully the fix will be available shortly in a production release.

I hope this helps. Until next time.....