Skip to main content

Scaling Unit - Why we should not have a master unit

In the last post we talk about the base concept of a Scale Units. In today post we will go further and talk about why we can end up with different types of scale units based on application needs.
Until now we defined a scale unit as a group of resources that are grouped together to server a specific number of clients.
This will work with success for application that can have scale units that are 100% independent and are not managed by an authority.
Let’s imagine the following scenario:
We need to create an application that pushes the same binary content to clients when an administrator decide.

It is pretty simple to define a scale unit that manage a X number of clients. This scale unit will contain the content that needs to be pushed to clients replicated as many as time needs to satisfy the SLA required by each client connected to that scale unit.

But there are some steps that need to be done before a release of a binary content can be done. For example we need to be sure that the binary content was downloaded with success by all scaling units. Also, we need to be able to control from only one point the release of a specific binary content.
If all this are not enough, when a client needs to be registered into the system, he will know an initial registration endpoint that will need to help him to find the scaling unit that will manage him.
Based on this requirement, we can already identify another type of scaling unit that will need to handle this functionality. Let’s call this scaling unit “Global Unit” – GU. The scaling units that handle our clients can be called “Client Unit” – CU.
The Global Unit will be aware all the time about the rest of the Client Units and will need to manage and communicate with each scaling unit.

This approach can be risky, because Global Unit can become a bottle neck. In case something goes wrong on it we will not be able anymore to trigger a global push of a binary content. Scaling a Global Unit is not an easy job, because it is the master controlling unit over the rest of our scaling units. We can have an Active-Active or Active-Passive failover mechanism at Global Unit, but it will be a hard thing to do because we’ll need to replicate content and storage between Active and the Passive (Active) unit (We will talk more about Active-Active and Active-Passive topic in another posts).
This is a risk that we need to accept if we will go on an approach with a Global Unit that manage and control Client Units.
It is important to know that not all the systems needs a Global Unit that control different commands over all Client Units - like a binary content release or redirect clients at the first handshake to their Client Unit. 
In general I would try to make everything that is possible to avoid having a Global Unit.

Our main scope should be to define scaling units that don’t have any kind of contact or shared resources between them or a 'master' unit. In this way scaling, management and control can be made clean and easily.

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