Skip to main content

Azure Data Lake and Replication mechanism cross Azure Regions

Context
Let’s imagine that we are working in an automotive company that collects telemetric data from their cars. A part of this data needs to be processed, stored and managed.
To be able to store data now and use it later on in time, you decided to go with Azure Data Lake, that is not limited on how much data you can store and allows you to plug any kind of processing system.
Requirements
After the architecture audit, because of legal constraints you are required to have a resiliency policy for disaster recovery. Even if in the same Azure Region there are 3 copies of data that are generated by Azure Data Lake, the legal constraints require you to have a resiliency policy.
Problem
Azure Data Lake makes 3 copies of data in the same Azure Region, but there is no support to replicated or backup content in a different Azure Region. You will need to define you own mechanism for this.

Available Solutions
We could fine many ways of doing this. There are 2-3 mechanism to do replication of Azure Data Lake cross region that is cost efficient and also not too complex from implementation point of view.
Azure Data Factory
One of the basic features of Azure Data Factory is support for copy activities. This is a generic service that allows us to move data from different types of storages. From binary content point of view, the current copy activities supported by Azure Data Factory offers us speeds up to 1 GBps for Azure Blob Storage and Azure Data Lake.
AdlCopy
This command can be used to move data from one Azure Data Lake to another. The command can be used with success as long as we integrate it in Azure Automation. This is required, because the command cannot run by its own. Azure Automation is the shell from where we run the command.

Downside of this solutions
This solution works well for small amount of data and when we don’t care about the deltas. If we can identify deltas by our self or we don’t care about it and we want to replicate everything than the previous two solutions are perfect.
If the process of delta identifications is too complex that we should go with Distcp command

Distcp
This can be used in combination with HDInsight to copy content from one storage to another. In comparison with the previous ones, it enables us to copy only the delta. Beside this, this command runs on top of a HDInsight (Hadoop) cluster, allowing it to scale inside the cluster based on how much traffic needs to be moved.

What’s suite our need?
Telemetric data collected from cars are grouped together per car and day and never updated. This means that we can identify easily what files need to be moved. Each car has a virtual folder with his own telemetric data. A new file is created each day. When a new car is sold, a new virtual folder is created.

This allows us to identify easily what files needs to be replicated and we don’t need to care about deltas, because we run the replication job only one time per day.
Taking this into account, Azure Data Factory is a good candidate.

Conclusion
As can see, there are multiple ways to do replication between different instances of Azure Data Factory. Even if Distcp is the best one, you need to take into account that you need to spin HDInsight clusters, that are more expensive that AdlCopy or Azure Data Factory.
When you handle storages where content is updated or copies rarely, then a solution based on Distcp might be too expensive and complex.

Comments

Popular posts from this blog

Windows Docker Containers can make WIN32 API calls, use COM and ASP.NET WebForms

After the last post , I received two interesting questions related to Docker and Windows. People were interested if we do Win32 API calls from a Docker container and if there is support for COM. WIN32 Support To test calls to WIN32 API, let’s try to populate SYSTEM_INFO class. [StructLayout(LayoutKind.Sequential)] public struct SYSTEM_INFO { public uint dwOemId; public uint dwPageSize; public uint lpMinimumApplicationAddress; public uint lpMaximumApplicationAddress; public uint dwActiveProcessorMask; public uint dwNumberOfProcessors; public uint dwProcessorType; public uint dwAllocationGranularity; public uint dwProcessorLevel; public uint dwProcessorRevision; } ... [DllImport("kernel32")] static extern void GetSystemInfo(ref SYSTEM_INFO pSI); ... SYSTEM_INFO pSI = new SYSTEM_INFO(

Azure AD and AWS Cognito side-by-side

In the last few weeks, I was involved in multiple opportunities on Microsoft Azure and Amazon, where we had to analyse AWS Cognito, Azure AD and other solutions that are available on the market. I decided to consolidate in one post all features and differences that I identified for both of them that we should need to take into account. Take into account that Azure AD is an identity and access management services well integrated with Microsoft stack. In comparison, AWS Cognito is just a user sign-up, sign-in and access control and nothing more. The focus is not on the main features, is more on small things that can make a difference when you want to decide where we want to store and manage our users.  This information might be useful in the future when we need to decide where we want to keep and manage our users.  Feature Azure AD (B2C, B2C) AWS Cognito Access token lifetime Default 1h – the value is configurable 1h – cannot be modified

What to do when you hit the throughput limits of Azure Storage (Blobs)

In this post we will talk about how we can detect when we hit a throughput limit of Azure Storage and what we can do in that moment. Context If we take a look on Scalability Targets of Azure Storage ( https://azure.microsoft.com/en-us/documentation/articles/storage-scalability-targets/ ) we will observe that the limits are prety high. But, based on our business logic we can end up at this limits. If you create a system that is hitted by a high number of device, you can hit easily the total number of requests rate that can be done on a Storage Account. This limits on Azure is 20.000 IOPS (entities or messages per second) where (and this is very important) the size of the request is 1KB. Normally, if you make a load tests where 20.000 clients will hit different blobs storages from the same Azure Storage Account, this limits can be reached. How we can detect this problem? From client, we can detect that this limits was reached based on the HTTP error code that is returned by HTTP