Azure Create Search Service and Scale Replicas and Partitions

In one of my previous articles AZURE SEARCH AS A SERVICE we have learned about the following related to Azure Search.

• What is Azure Search?
• What are Azure Search advantages?
• What are Azure Search features?
• Process of creating and using the Search function
• What data sources are supports?

In this article, let’s discuss about the 1^st step Create Azure Search – highlighted in the below diagram using the Azure Management portal.

Navigate to the Azure Management portal and select “Azure Search” as shown below.

Clicking on the Azure Search in the above screen capture, you will be taken to the following New Search Service blade.

You need to provide the following details.

Field Name	Description
URL	A unique URL will be created based on the value that provide in this textbox. In this example, I have provided “smartazuresearch”. So, the URL will be https://smartazuresearch.search.windows.net . This is the URL that you would be using in the API’s that you use for pushing the data to Search indexes. You will learn more about this in the upcoming articles. Please note that you cannot change this URL once you create it.
Subscription	Select the subscription in which you would like to create the service.
Resource Group	Choose the existing resource group in which your related applications are created. It’s a good practice to have all the services of a given project’s environment (Production / Staging / Dev) in the same Resource Group
Location	Select the Data Center location where you would like to have Azure provision the Search Service. Please make sure that all the applications that consume your Search Service are also hosted in the same location to avoid latency issues.
Pricing Tier	In this example, I have chosen Standard – S1 tier. Unlike App Service Plan’s tiers, you CANNOT do an in-place updating of the pricing tier of the Search Service. In case if you would to scale-up from Free to Standard, you need to re-create everything and migrate the index data (we will discuss more about indexes in the upcoming articles).

Once you are ready with all the details, click on Create to create the Azure Search service. Below is the screen capture of the service that we have created.

Below are the features that different tiers offer as of this writing.

	Free	Basic	Standard – S	Standard -S2	Standard – S3	S3 High-Density
Indexes	3	5	50	200	200	1000
Documents	10K	1M	15 M	60M	120 M	200 M
Storage (GB)	50 MB	2 GB	25 GB / Partition	100 GB / Partition	200 GB / Partition	200 GB / Partition
Infra	Shared	Dedicated	Dedicated	Dedicated	Dedicated	Dedicated
Scaling (Search Units)	None	3	36	36	36	36
Replicas (Load Balancing)	None	3	12	12	12	12
Partitions	None	None	12	12	12	3
High Density Enabled	No	No	No	No	NO	Yes

We will learn more about all the above specified features in-detail in the upcoming articles.

Scaling – Azure Search could be scaled in the following two broader ways.

• Replicas – Replicas distribute workloads across the service. You need 2 or more replicas to support high availability (applies to Basic and Standard tiers only).
• Partitions – Partitions allow for scaling of document counts as well as faster data ingestion by spanning your index over multiple Azure Search Units (applies to Standard tiers only).

Navigate to Settings blade and click on Scale as shown below.

By default, each service that you create has 1 replicate and 1 partition. We can scale up / down using the Scale blade as shown below. In the above screen capture, I scaled up the replicas to 2.

In production environment, you should look at having a minimum of 3 replicas to avail 99.9% availability.

Summary: In this article, we have learned the following.

• How to create an Azure Search service using Azure Management Portal.
• How to scale-up replicas and partitions to increase the availability of the Search Service.

Hope you enjoyed reading the article. Your feedback is highly appreciated.

Do you like this article? If you want to get more updates about these kind of articles, you can join my Learning Groups

WhatsApp

Telegram