Issues while deploying Elastic Search from Azure Portal

Hi Team,

I am trying to deploy Elastic Search(Self- Managed) from Azure portal.

However the issue I am facing is not able to select the virtual network as seen below:

The virtual network(SUB_VNET_29_0) initially had one subnet with /26 address space.

However after going through Elastic Search documents, came to know that constraint to

deploy Elastic Search from Marketplace UI is to have /25 address space. Hence we have added a

new subnet to the same virtual network(SUB_VNET_29_0) with /25 address space. But still the

Elastic Search deployment from Azure UI is not listing the virtual network though the newly added

subnet has 128 IP address space.

Any ideas on this please?

Thanks & Regards,

You may get an answer here from another Azure user but this really is an Azure question - Microsoft's support service and forums may get you a faster answer.

As @Emanuil says, this does sound like an issue in the portal UI with the Microsoft.Network.VirtualNetworkCombo UI element. The rules for the network for the Elasticsearch solution template are

In addition to having a subnet min address prefix size of /25, there must be a min available number of addresses of 3 and addresses must be contiguous.

The min address prefix size of /25 is just an Azure portal UI constraint and is not a constraint of the underlying Azure Resource Manager (ARM) template that the Marketplace offering calls behind the scenes. If that's an option in the meantime, take a look at the getting started docs for deploying the ARM template using Azure CLI 2.0 or Azure PowerShell.

Hi Russ,

Thanks for the detailed explanation. :slight_smile:

In my case though the newly added subnet has(/25) address space, the virtual network had (/26)
address space, meaning it is not satisfying the network constraint of Elastic Search.
Will add a new virtual network of (/24) address space and subnet of (/25) address space and see if it works from the UI.

Coming to the deployment of Elastic Search using the ARM template.

Followed the link you have provided and however not able to deploy it for the latest version(7.6.2). Throws error as: " Unable to download the deployment content from the..."
Also I noticed that we are not able to open the 7.6.2 json from browser as well.

Next i tried deploying 6.6.1 by passing the existing network parameters , not sure why it is not picking up the parameters I have passed. The code is still trying deploying with new network.

Kindly let me know if the network parameters syntax is correct or not.

Thanks & Regards,

Please don't post images of text as they are hard to read, may not display correctly for everyone, and are not searchable.

Instead, paste the text and format it with </> icon or pairs of triple backticks (```), and check the preview window to make sure it's properly formatted before posting it.

This would mean that the subnet address space is larger than the virtual network address space, which I hope Azure would not allow you to do. If you create a vnet with at least /24 and a subnet with at least /25. it'll work.

What's the exact error message? Can you provide the steps that you went through?

I don't understand what you mean. Can you clarify please?

6.6.1 is not a supported version with the template. If deploying from the master branch, the versions supported are:

How are you asserting this?

Hi Russ,

Thanks a lot for your time in going through this and replying back. Appreciate your help. :slight_smile:

As you have said , after adding a new virtual network of (/24) address space and a new subnet to it with (/25) address space, I am able to select the network while deploying the ELK in Azure. The new network we have added is and as per our policy we dont allow 10 series IP's. However while going through the sequence of steps, and selecting the Logstash and Kibana to be installed, at the end it is failing to create because of the policy violation errors, means it is not picking up the network we have allocated.

Any ideas on this?

Thanks & Regards,

Does the installation of Kibana and Logstash in Azure need the public IP's?

The template dynamically assigns IP addresses, so is sounds to me like the Azure infrastructure does not take this policy into account when dynamically assigning addresses. Sounds like you may need to exclude the deployment from the policy.

Kibana is configured with a public IP address, Logstash is not. This is not configurable in the template. Deploying a standard SKU loadbalancer also assigns public IP resources to each virtual machine, which is needed in this scenario in order to install the Elastic Stack.

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.