Skip to main content

Azure Batch Service - Overview (part 1)

In the last few days a new Azure service catch my eyes. Unfortunately only today I had time to write about it.
Azure Batch Service is the today star. Why? Because enables us to run tasks over Azure infrastructure without having to think about scaling, parallel and performance problems. You only need to write the batch logic, upload to Azure and everything else will be managed by Azure Batch Service.
On top this, we have a scheduling and auto scaling mechanism that enable us to have more granular control to compute resources. The API give us the possibility to publish and run batch applications without thinking about resources, service management and so on. Also, Batch Service provide us full access to configuration, giving us the possibility to manage all the resources and task execution pipeline. In this case you need to manage how content is moved, persisted and so on.

What you can run? 
There are two types of formats that can be used in this moment. An classic .exe file that represents you batch application or a batch application (that is represented by a collection of assemblies, that implements specific interfaces – we will talk about this subject later on).

What is a TVM?
Task Virtual Machine (TVM) is an Azure VM where your tasks will run. All your workload created by your tasks will run on one or more dedicated TVM. User has the possibility to define a pool of TVM that will be managed by Azure.
When you create a pool of TVM you have the possibility to specify different attributes like size, OS, pool size, scaling policy, certificates and many more. In one world you have full control of what kind of TVM are used for batch tasks. And yes, each TVM has his own IP and name.

What is a work item?
Work item is the template that specify how the application will run on TVM. It is the interfaced used by developers to control how a batch application runs.

What is a job?
A job is an instance of a work items that runs on TVM.

What is a task?
A task is the logic that is executed on TVM. One or more tasks define a job.

Do I need to manage all of them (work items, jobs and tasks)?
No. If you are adding an existing application and you only want to run in over Batch infrastructure, than the jobs, splitting into tasks, merging result and many more will be full managed by Batch Services.
On the other hand, if you want a full control of Batch, that you will need to manage and implement all of them.

What is a Merge Task?
It is the step when the results from each task is merged by the job to create the final result.

What is a file?
File represents the input data used by jobs. Based on your logic a task can use one or more files.

What is a directory?
Directory is the location under each task that is used to run the program and other temporary files. All file and directories under a job, can be accessed by the tasks of that job.

Where I need to store all this files?
All the files needs to be stored under the storage account of the subscription that is used by Batch Service.

How I trigger a batch action?
In general a batch by submitting a job. For input data, the files will be used that are related to that job. This can be done over REST API very easily.

Can I monitor a job?
Yes, when a job runs, you can monitor it (progress status) and access the output data at the end.

In the next post we will look over the code and discover how we can work with Azure Batch Service.

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