Richard Beeston, CBA®’s Post

The Save operation was implemented in Duva(https://lnkd.in/gEjkJ26e) based on the design outlined in the following post. This marks another milestone. Yet, it leaves something to be desired. 1) Optimization on buffering to reduce syscall 2) Graceful shutdown that stands Sigterm Guess it's the right timing to consider a system-level graceful shutdown. Why hasn't it been there? Everything was done in memory, and the assumption was that there was no save, no persistence, and no need for recovery as nothing needed to be recovered.

Application Architecture: The Importance of Making the Right Choice Early On When it comes to application design, one of the most critical decisions is choosing the right architectural approach. Mistakes at this stage can cost time, resources, and performance down the line. 🔑 Key Points to Consider: 1. Modularity and Extensibility – choose an architecture that is easy to adapt to changes and allows new features to be added without rewriting code. 2. Scalability – how will your architecture handle increased load? Both vertical and horizontal scalability should be built into the design. 3. Performance and Reliability – the structure of your application should guarantee minimal latency and resilience to failures. 4. Error Handling and Fault Tolerance – a well-thought-out error handling system not only improves the user experience but also reduces the risk of a complete system crash. 5. Security – ensuring security at every level of architecture is a priority, especially in today’s world of cyberattacks. 6. Technology Stack – choose a stack that not only solves today’s problems but will remain relevant in the coming years. 💡 Tip: Sometimes less is more. Don’t overcomplicate your architecture with complex solutions when simple design patterns can do the job.

You can build it yourself, but what will it take?? 🤔 The Build vs Buy question is one we discuss with potential customers all the time. It’s a no-brainer - you’re great engineers - an authorization system is something you CAN build in-house. We know you can. That’s how all companies have done it so far, and it's how many companies do it today. But the real question is - should you? Building an authorization layer from scratch takes time and effort from your development team. This time and effort you could spend on developing the core, unique features of your product. But that’s a given - what many people don’t think about is the upkeep and maintenance of such a system over time. * How do you handle performance? it’s a lot of work to get to 1-6ms per authorization query at p95 - naive solutions give you around 300ms - and that basically kills your app. * What happens when you need to grow to millions of identities and/or resources? Your app is (hopefully) going to grow. Will your authorization be able to adapt to major application changes? or will you have to build the entire thing from scratch? * Anyone can do basic ACLs or RBAC - but what happens when you have more complex requirements? Multi-tenancy, dynamic and custom roles, time-based access, location-based, org chart-based, PII filtering - these all might be just around the corner. * How do you make it usable for everyone? Building CI/CD, UIs for non-technical users, IaC for policy as code, access control for the access control, auditing, etc. I can keep going. If you have an infinite amount of development resources to spend on this, that’s one thing (Also let me know how you achieved that). Otherwise, you must ask yourself these questions. If you are part of a small/medium-sized organization, you must ask yourself these questions. And once you do, there’s a good chance you’ll realize: * It will be expensive to build * It will be expensive to maintain * It’s extremely hard to predict a system that will be future-proof from day 1 Bottom line - it’s very difficult to know in advance what solving this problem would take. Using the aid of an external provider can help solve this problem for you, and, perhaps even more importantly, attach a quantifiable sum to it. As always thank you for the meme Daniel Bass

𝗘𝘅𝗽𝗹𝗼𝗿𝗶𝗻𝗴 𝗥𝗘𝗦𝗧 𝗔𝗣𝗜 𝗶𝗻 𝗮 𝗡𝘂𝘁𝘀𝗵𝗲𝗹𝗹 ! 𝟭. 𝗪𝗵𝗮𝘁'𝘀 𝗮 𝗥𝗘𝗦𝗧 𝗔𝗣𝗜? - It's like a messenger that lets different apps or websites talk to each other. - It uses a clear language called HTTP to send and receive messages. 𝟮. 𝗛𝗼𝘄 𝗱𝗼𝗲𝘀 𝗶𝘁 𝘄𝗼𝗿𝗸? - 𝙍𝙚𝙦𝙪𝙚𝙨𝙩: ∙ A client (like a web browser or app) sends a request to the server. ∙ It uses different HTTP methods like GET (to fetch data), POST (to create data), PUT (to update data), or DELETE (to remove data). - 𝙍𝙚𝙨𝙥𝙤𝙣𝙨𝙚: ∙ The server sends back a response with a status code (like 200 for success or 404 for not found). ∙ It also includes the requested data or any error messages. 𝟯. 𝗥𝗘𝗦𝗧 𝗔𝗣𝗜 𝗖𝗼𝗻𝘀𝘁𝗿𝗮𝗶𝗻𝘁𝘀: - 𝘾𝙡𝙞𝙚𝙣𝙩-𝙎𝙚𝙧𝙫𝙚𝙧 𝘼𝙧𝙘𝙝𝙞𝙩𝙚𝙘𝙩𝙪𝙧𝙚: Decouple the user interface concerns from server logic. - 𝙐𝙣𝙞𝙛𝙤𝙧𝙢 𝙄𝙣𝙩𝙚𝙧𝙛𝙖𝙘𝙚: Foster a consistent and intuitive communication interface. - 𝙇𝙖𝙮𝙚𝙧𝙞𝙣𝙜: Enable modular design with layered architecture. - 𝙎𝙩𝙖𝙩𝙚𝙡𝙚𝙨𝙨𝙣𝙚𝙨𝙨: Each request from a client contains all the information needed to understand and fulfil the request. – 𝘾𝙖𝙘𝙝𝙚-𝙖𝙗𝙞𝙡𝙞𝙩𝙮: Optimise performance by allowing responses to be cacheable. ---------------------------------------------------------------- follow: Jagdish Saini Credit: Hina Arora

Signs Your Software Architecture Needs A CPR 1. Monitoring gives more errors than a GPS navigation system in a corn maze. 2. Developers need more time to start their testing environment than it takes for a sloth to finish a marathon snack break. 3. Fixing bugs introduced by recent deployments feels like trying to untangle a spaghetti knot blindfolded. 4. All tests refuse to play nice on the developer's machine, staging a rebellion akin to toddlers at bedtime. 5. Injecting failures into the system feels like playing Russian roulette with a Nerf gun. 6. Unit tests, numbering in the thousands, take longer to run than a tortoise crossing the finish line. 7. The domain model is about as reliable as a politician's campaign promise. 😄 8. Flaky tests make developers question reality more than a conspiracy theorist on a caffeine bender. 9. People are more afraid of making mistakes than a tightrope walker with vertigo. 10. DevOps folks can't test changes on their local machines without breaking a sweat, feeling more uncertain than a squirrel in a game of Frogger. 11. Someone always forgets to update the documentation manually, leading to more confusion than a game of telephone with a group of gossiping squirrels. 12. The testing pyramid looks more like a Jenga tower after a few too many rounds of drinks. 13. Monitoring is about as effective as trying to stop a leaky dam with a bucket full of holes. 14. Business rules are scattered like socks after a laundry mishap, desperately needing a dedicated home base like a lost puppy.

An Overview of Devin's Agentic Architecture 🤖 Here's how it works—Devin continuously operates in a loop, refining the code, browsing for data, and executing tasks until everything is done. It’s a dynamic agent that integrates web browsing, coding, and execution tools in one simple to use chat interface. Here’s Devin’s process in a nutshell: ⎆ Task Start: You assign a task to Devin. It could be a code-related task or something that needs data from the web. ⎆ Loop Logic: Devin enters its decision-making loop: ⎆ If it needs to write or edit code, it uses Cursor. ⎆ If it needs data from the web, it uses Multion to browse the web. ⎆ Conditional Execution: Once the code is prepared, Devin decides whether to execute it or continue refining. If execution is required, it uses e2b dev to run the code and get real-time results. ⎆Looping: After each action, Devin goes back to decide what to do next. It will keep running this loop—gathering data, editing code, and executing—until the task is fully completed. ⎆ This N-loop system ensures that the agent continues to iterate until the task meets all requirements. Want to know more about how these agents can improve your workflow? Like, share, and follow to receive more agent content!

An Overview of Devin's Agentic Architecture 🤖 Here's how it works—Devin continuously operates in a loop, refining the code, browsing for data, and executing tasks until everything is done. It’s a dynamic agent that integrates web browsing, coding, and execution tools in one simple to use chat interface. Here’s Devin’s process in a nutshell: ⎆ Task Start: You assign a task to Devin. It could be a code-related task or something that needs data from the web. ⎆ Loop Logic: Devin enters its decision-making loop: ⎆ If it needs to write or edit code, it uses Cursor. ⎆ If it needs data from the web, it uses Multion to browse the web. ⎆ Conditional Execution: Once the code is prepared, Devin decides whether to execute it or continue refining. If execution is required, it uses e2b dev to run the code and get real-time results. ⎆Looping: After each action, Devin goes back to decide what to do next. It will keep running this loop—gathering data, editing code, and executing—until the task is fully completed. ⎆ This N-loop system ensures that the agent continues to iterate until the task meets all requirements. Want to know more about how these agents can improve your workflow? Like, share, and follow to receive more agent content!

An Overview of Devin's Agentic Architecture 🤖 Here's how it works—Devin continuously operates in a loop, refining the code, browsing for data, and executing tasks until everything is done. It’s a dynamic agent that integrates web browsing, coding, and execution tools in one simple to use chat interface. Here’s Devin’s process in a nutshell: ⎆ Task Start: You assign a task to Devin. It could be a code-related task or something that needs data from the web. ⎆ Loop Logic: Devin enters its decision-making loop: ⎆ If it needs to write or edit code, it uses Cursor. ⎆ If it needs data from the web, it uses Multion to browse the web. ⎆ Conditional Execution: Once the code is prepared, Devin decides whether to execute it or continue refining. If execution is required, it uses e2b dev to run the code and get real-time results. ⎆Looping: After each action, Devin goes back to decide what to do next. It will keep running this loop—gathering data, editing code, and executing—until the task is fully completed. ⎆ This N-loop system ensures that the agent continues to iterate until the task meets all requirements. Want to know more about how these agents can improve your workflow? Like, share, and follow to receive more agent content!

Design Pattern - Best Practice From MFV's SCI Project Ever struggled with designing software that's both efficient and maintainable? Design Patterns are the solution you need! Why? Here are 3 reasons • Common Language for Developers: They help you communicate and share ideas more effectively. • Reusability and Extensibility: Write maintainable and extendable code without affecting other parts of the system. • Increased Testability: Easily test each function without impacting the entire system. Click the thumbnail for more. Follow Linkedin for more updates of MFV's Best practices. Thank you Nani for the topic.

𝗘𝘅𝗽𝗹𝗼𝗿𝗶𝗻𝗴 𝗥𝗘𝗦𝗧 𝗔𝗣𝗜 𝗶𝗻 𝗮 𝗡𝘂𝘁𝘀𝗵𝗲𝗹𝗹 ! 𝟭. 𝗪𝗵𝗮𝘁'𝘀 𝗮 𝗥𝗘𝗦𝗧 𝗔𝗣𝗜? - It's like a messenger that lets different apps or websites talk to each other. - It uses a clear language called HTTP to send and receive messages. 𝟮. 𝗛𝗼𝘄 𝗱𝗼𝗲𝘀 𝗶𝘁 𝘄𝗼𝗿𝗸? - 𝙍𝙚𝙦𝙪𝙚𝙨𝙩: ∙ A client (like a web browser or app) sends a request to the server. ∙ It uses different HTTP methods like GET (to fetch data), POST (to create data), PUT (to update data), or DELETE (to remove data). - 𝙍𝙚𝙨𝙥𝙤𝙣𝙨𝙚: ∙ The server sends back a response with a status code (like 200 for success or 404 for not found). ∙ It also includes the requested data or any error messages. 𝟯. 𝗥𝗘𝗦𝗧 𝗔𝗣𝗜 𝗖𝗼𝗻𝘀𝘁𝗿𝗮𝗶𝗻𝘁𝘀: - 𝘾𝙡𝙞𝙚𝙣𝙩-𝙎𝙚𝙧𝙫𝙚𝙧 𝘼𝙧𝙘𝙝𝙞𝙩𝙚𝙘𝙩𝙪𝙧𝙚: Decouple the user interface concerns from server logic. - 𝙐𝙣𝙞𝙛𝙤𝙧𝙢 𝙄𝙣𝙩𝙚𝙧𝙛𝙖𝙘𝙚: Foster a consistent and intuitive communication interface. - 𝙇𝙖𝙮𝙚𝙧𝙞𝙣𝙜: Enable modular design with layered architecture. - 𝙎𝙩𝙖𝙩𝙚𝙡𝙚𝙨𝙨𝙣𝙚𝙨𝙨: Each request from a client contains all the information needed to understand and fulfil the request. – 𝘾𝙖𝙘𝙝𝙚-𝙖𝙗𝙞𝙡𝙞𝙩𝙮: Optimise performance by allowing responses to be cacheable. ---------------------------------------------------------------- Post credits @HINA ARORA 📌 Save this post for later and follow Amit Sharma for more insightful information.