h a l f b a k e r yIt's not a thing. It will be a thing.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
Please log in.
Before you can vote, you need to register.
Please log in or create an account.
|
Architecture scaling plan program
Computers are usually slow due to their architecture. Architecture is set in stone at design time when software is designed. This idea is that we write a program to generate the architecture and evaluate costs based on the plan. | |
This idea hinges on the idea that there is an drawable model of software architecture that can be stored as data, then calculations can be applied to the plan to work out how much it shall cost.
We write a program that has a slider that goes infinitely far into the distance.
You can drag the slider
and the architecture gets more advanced and expensive and scalable.
Architecture includes things such as object graph relationships, bounded contexts, responsibilities, push/pull, async/sync, quantity of data, latency vs throughput, number of concurrent users etc, traffic, amount of data, retention policies, etc
Computer problems could be solved from just write shell script or Python script, all the way up to a program on server, perhaps in Java, then separate database box, then multiple web servers, then load balancers in the Cloud (AWS/GCP/Azure) and so on, then clusters and Hadoop and ETL and data warehouses such as Snowflake.
It gets more and more complicated.
progression of scaling
http://highscalabil...ers-on-amazons.html
A Beginner's Guide To Scaling To 11 Million+ Users On Amazon's AWS [chronological, May 14 2023]
[link]
|
| |
So, is this a tool to design hardware optimised for specific software? |
|
| |
I was not expecting strong AI to be necessary. |
|
| |
Many architectures can be drawn as simple diagram, bottlenecks exist in the diagram and the transformation of the diagram can be mechanical. |
|
| |
For example, if you start with one server, you might have a rule that you split it in half and put 2. |
|
| |
I have linked a highscalability article of the progression required to scale to 11 million users. |
|
| |
Please have a look at that article and imagine a program that implements that progression automatically. |
|
| |
pertinax, this is for software architecture. |
|
| |
So load balancers, servers, web servers, database servers etc. |
|
| |
Thank you for that clarification. I'm still thinking about whether this requires strong AI, AI or just I. |
|
| |