Seferlis ch3

• Core Azure services include storage accounts, which are used for storing data that is used in cognitive services solutions.
• When considering costs, it's important to find a balance between services and costs that meets your business requirements.
• Service-level agreements (SLAs) do not guarantee service, they simply provide a mechanism to recover costs that are incurred while a service is down.
• If a service goes down for longer than the stated SLA over a month, you can get service credit back from Microsoft for that lost time.
• It's advisable to work with Microsoft directly or request Microsoft-certified experts to review your proposed architecture to ensure guaranteed uptime of your services.
• Performance is considered when creating a storage account to determine whether the low latency performance of the premium tier is needed.
• A storage account name is the name of the storage account and cannot be changed, so choose carefully.
• Hierarchical namespace is an optional setting under the Advanced tab and should be considered based on how the data lake is set up.
• A resource group is a logical container used to contain all the Azure services used for the workload.
• Tags are optional and help organizations search for resources and build reports around services that can help with billing segmentation.
• A region is selected to determine where the services will be hosted.
• Redundancy is chosen while deploying a storage account, depending on the environment it is being deployed for.
• As the services evolve, more features and capabilities are added, so be sure to watch the public roadmaps for upcoming features.
• SLA, HA, DR, and other resiliency options are determined based on Ocean Smart's specific deployment needs.
• Data movement tools to help move data to Azure efficiently for specific requirements are provided in this chapter.
• From the next chapter onward, the book will explore the benefits of building a knowledge mining solution with Azure Cognitive Search.
• Modeling costs for solutions are considered as the book progresses and builds solutions for specified use cases.
• This can also include new generations of products, new regions for deployment, and price changes that affect these services.
• When building cognitive services solutions, one of the most critical components is clean, trusted data that will be used as a basis for all the predictions or observations that will be made as part of your solution.
• The data that's being used is typically stored in a storage account and used for downstream activities that are performed to build the complete solution.
• When deploying a storage account, it's important to consider services that are both required and optional.
• Required services when deploying a storage account include the subscription, which is the level where all the resources are created and rolled up to for billing purposes.
• Different organizations have different policies as to how subscriptions are set up, and if you do not have one yet, consider working with finance and accounting to ensure the structure of your subscriptions adheres to common payment processes.
• Optional services when deploying a storage account include the location, which is used to determine the physical location of the storage account.
• The other optional service when deploying a storage account is the encryption type, which determines the level of encryption for the storage account.
• The Curated Zone, sometimes known as the finished or trusted zone, is the area where data will be trusted as clean and complete for the final phase before being used in a data model training or similar operation.
• The developer can build zones for moving the data through the work flow.
• The Raw Zone, also known as a landing zone, is where the data will land into the storage account in Azure, where it is intended to stay in its raw form and archived over time.
• The best way to get started with pricing is to visit the Azure pricing calculator to build an estimate: https://azure.microsoft.com/en-us/pricing/calculator/
• The zones of ADLS Gen 2 can be laid out with folder and file hierarchies within each zone.
• Data storage in Azure is always stored across three locations for redundancy purposes, but depending on your regional requirements, storage can be replicated in one of several ways, such as LRS or ZRS.
• The Staging Zone is designated to be transient in that redundancy and high availability are not necessary.
• Azure Storage costs are determined by numerous factors, such as region, type, performance tier, storage account type, access tier, and redundancy.
• In the Staging Zone, data can be manipulated for aggregation, cleansing, enhancement, and more.
• When building cognitive services solutions, it's important to consider the costs and services that meet your business requirements.
• The access tier helps the creator specify how the data will be stored concerning how frequently it will need to be accessed.
• Azure Storage allows for two performance tiers: Standard and Premium.
• Redundancy options that are available determine the resiliency of the data that has been stored.
• The three current access tier options are Hot, Cool, and Archive.
• Azure VNet offers several capabilities that are required for workloads to be deployed successfully, many of which align with the same core requirements of networking in the past.