다음을 통해 공유

Azure Storage: Overview Of The Azure Storage Services (Part 2)

  • 아티클

 

Series

This article is a part of a series. 

Check the first part here: Azure Storage: Overview Of The Azure Storage Services (Part 1).

Introduction

In the Part 2 of this series post of Azure Storage Service, we will talk about the Azure Storage Services and their details, also we try to figure out how we can calculate the cost of these and the tools that we can manage Azure Storage Services.

Azure Storage Services

There are four Azure Services available:

  • Blob
  • Files
  • Tables
  • Queues

Azure Blob

The word Blob means "Binary Large Object". Using this service we are able to store a large amount of unstructured data. For example, we can store types of files like images, videos, text files, log files, vhd, iso files, etc.

Type Of Blobs

There are three types of blobs :

  • Page Blobs: Page blobs are used for frequent read-write operations. Every Page Blob consists of pages of 512 bytes each. The maximum size for a page blob is 8 TB. Page Blobs are used in Azure Virtual Machines, OS and Data Disks.
  • Block Blobs: Block blobs are used for images, video files, etc. Each Block Blob is composed from blocks of data which can be managed individually. Each Block Blob has a maximum size of 4.7 TB.
  • Append Blobs: Append Blobs are similar to Block Blobs, with the difference that are the best choice for append operations like Log Files.

 

Limitations

At the table below we can see the Azure Blob Scale Targets:

Resource  Target
Max size of the single blob container  Same as max storage account capacity
Max number of blocks in a block blob or append blob 50.000 blocks
Max size of a block in a block blob 100 MB 
Max size of a block blob 50.000 x 100 MB (approx. 4.75 TB)
Max size of a block in an append blob 4 MB
Max size of an append blob 50.000 x 4 MB (approx. 195 GB)
Max size of a page blob 8 TB
Max number of stored access policies per blob container 5
Targe throughput for a single blob Up to 60 MB  per second, or up to 500 requests per second

Default Endpoint: http://mystorageaccount.blob.core.windows.net

↑ Back to top

Azure Files

Azure Files supports the Server Message Block protocol standard (SMB) and gives the capability to create a file share in Azure that can be mounted to an on-premises workstation as a mapped network drive. The Azure Files can be accessed from multiple points.

 

Limitations

At the table below we can see the Azure File Share Scale Targets.  

Resource Standard file shares  Premium file shares (preview)
Minimum size of a file share (no minimum; pay as you go) 100 GB
Max size of a file share 5 TB 5 TB
Max size of a file in a file share 1 TB 1 TB
Max number of files in a file share No limit No limit
Max IOPS per share 1000 IOPS 5120 IOPS baseline / 5.360 IOPS with burst
Max number of stored access policies per file share 5 5
Target throughput for a single file share Up to 60 MB/sec Up to 612 MB/sec (provisioned)
Maximum open handles per file 2.000 open handles 2.0000 open handles
Maximum number of share snapshots 200 share snapshots 200 share snapshots

Default Endpoint: http://mystorageaccount.file.core.windows.net

↑ Back to top

Azure Tables 

Azure Tables is a solid cloud NoSQL storage which is massively scalable. Azure Tables can be accessed via REST API. It is a great solution for cases with a schemaless design. 

Limitations

At the table below we can see the Azure Tables Scale Targets.

Resources Target
Max size of a single table 500 TB
Max size of a table entity 1 MB
Max number of properties in a table entity 255 (including 3 system properties: PartitionKey, RowKey, and Timestamp)
Max number of stored access policies per table 5
Maximum request rate per storage account 20.000 transactions per second (assuming 1 KB entity size)
Target throughput for single table partition (1 KB entities) Up to 2000 entities per second

Default Endpoint: http://mystorageaccount.table.core.windows.net

 

Azure Queues 

The Azure Queues storage service is appropriate for storing and exchange messages using HTTP/HTTPS calls. This is a 'low latency with high throughput' system. A message size can be 64 KB and the maximum number of messages is defined from the storage account capacity.

 

Limitations

At the table below we can see the Azure Queues Scale Targets.

Resource Target
Max size of a single queue  500 TB
Max size of a message in a queue  64 KB
Max number of stored access policies per queue  5
Maximum request rate per storage account 20.000 messages per second assuming 1 KB message size
Target throughput for the single queue (1 KB messages) Up to 2000 messages per second

Default Endpoint: http://mystorageaccount.queue.core.windows.net

↑ Back to top

Azure Storage Cost 

In the following links, we can find all the useful information on how to calculate Azure Storage consumption.

Managed Disk

The formula: Managed Disk Cost = Fixed Cost (Per Disk Size) +  Operations Cost

The Example: Managed Disk Cost = S10 128GB (€4.97) + (200.000 Transaction units * €0.0003 Per unit) = €65.68

Unmanaged Disk

The formula: Unmanaged Disk Cost = (Storage Usage In GB * Cost Per GB) + Operations Cost

The Example: S10 128GB (€4.86) +  (200.000 Transaction units * €0.0005 Per unit) = €91.08 *

 

Note

*Unmanaged disks billing is based on the used storage space, the example supposes that all 128 GB are used.

The formulas apply on the Standard HDD and Standard SSD disks, while no transaction costs are applied for the Premium SSD disks.

The examples are calculated for Operations Cost = 200.000 Transaction units. Each Transaction unit is equal to 10.000 transactions.

↑ Back to top

↑ Back to top

Azure Storage Tools 

There are several tools (Microsoft or 3rd Party) that help us to manage Azure Storage Services.

Microsoft Tools

3rd Party Tools

↑ Back to top

Conclusion 

After reading this post series we have a good grasp of Azure Storage Services. We are in a position to know how important is to implement a strategy for the Azure Storage Services usage. Microsoft offers many options that can suit all business needs in terms of performance, data replication and budget, and each of them adheres to its respective service limits and benefits.

↑ Back to top

See Also 

 

↑ Back to top