In the Kubernetes world there are basically three cascading deletion flavours:– background (DEFAULT)– foreground– orphan Background This strategy deletes the resource itself and all their dependants recursively down immediately (doesn’t wait for dependants to be deleted first). And as it is the default cascading strategy you don’t need to specify it, for example: Foreground This …
HTTP provides (RFC 9112) a way for sending several chunks back by a server as a response to the request. Let’s see how to communication looks like. Communication A client sends a normal request, like for example: And a server responds in several chunks like: Response structure Each chunk contains at least two parts:– size …
How to send an HTTP response in chunks? Chunked Transfer Encoding Czytaj dalej »
Dropping a role seems to be a simple thing like: but it is not… In order to drop a role you need to ensure that there are no objects bound to this role. How to unbind objects from a role? Do it for every database! What is really important while unbinding objects from a role …
There are at least two not obvious class initialization mechanisms in Python, especially if your background is in different programming languages π Class field initialization If you assign a class value outside __init__ method like here (see class_field): You should get: So be careful – class fields initialized outside __init__ are initialized only once (when …
There are only 2Β³Β² = 4,294,967,296 IP possible addresses in version 4 (IPv4) and around 8,000,000,000 people in the world. So it means that only roughly half of the citizens can have one IP address. It’s a big limitation considering at least a few devices per one person. So there are two solutions for this …
What is the difference between 127.0.0.1 and 0.0.0.0 when you bind your server to one of them? Simple answer is – 127.0.0.1 never goes outside your machine, whereas 0.0.0.0 does. 0.0.0.0 So as depicted in the picture – when your server is bound to 0.0.0.0 it means that it listens to all interfaces (traffic comming …
In a distributed architectures (like microservices or event-driven architecture) sometimes a client request requires information from more than one service and this is called workflow management. There are primarly two types: orchestration and choreography. Orchestration Orchestration means that there is one service (conductor) which takes care of the workflow process. Pros of Orchestration Cons of …
Workflow management – orchestration versus choreography Czytaj dalej »
Here I’ve listed a few characteristics (with their descriptions, if needed) which can be used while thinking about an architecture of the system: Scalability Usually it is about horizontal scalability of independent services. Elasticity If a system can scale up or down automatically. Availability It is how much time it must be available. Some systems …
Layered architecture Architectural Characteristic Star Rating Maintainability β Testability β β Deployability β Simplicity β β β β β Evolvability β Performance β β β Scalability β Elasticity β Fault Tolerance β Overal Cost β Modular monolith architecture Architectural Characteristic Star Rating Maintainability β β β Testability β β β Deployability β β β Simplicity β β β β Evolvability β β β Performance β β β Scalability β Elasticity β Fault Tolerance β Overal …
Star rating of architectural styles [Head First – Software Architecture] Czytaj dalej »
CDN stands for Content Delivery Network and it’s idea is simple – instead of serving content from one server, let’s put multiple replicas (with the same content) in different geographical places. So thanks to this an application can gain multiple benefits:– faster loading time– reduced traffic on the main server– better reliability (if one server …
