RDS aurora serverless 15 read replicas/3AZ read replicas are not the same as Disaster Recovery Multi AZ is not the same as multi region By default, RDS is not multi AZ multi AZ cross regions does not exist for RDS, only for Aurora it exists aurora endpoint storage autoscaling RDS proxy <=> connection pool maintenance:…
amplify web or mobile app WAF geographic restriction, before ALB, IP filter, pattern, limit request Road 53 100 % availability, smart routing routing policies SNI: service name, multiple @ geolocalisation dns suppport, dns host names firewall manager waf shield advanced SG route 53 network firewall VPC (SG)
launch template (delete old version and launch new version) launch config is deprecated scaling policies target: metric step scheduled predictive use case: you want at least 4 servers : 6 servers on 3 AZ or 8 servers on 2AZ ? 6 servers on 3AZ, better for cost and for the goal termination hook placement group…
VPC Security group is stateful, NACL is stateless NACL needs inbound and outbound rules Default SG is not enought A public subnet does not imply a public IP Bastion Host
EC2 Not ServerLess install cloudwatch agent delete on termination NITRO 64 000 IOPS io (multi attach EBS) needs 5 IP (on the CIDR) .1, .2.3.3 .4 .255 recover EC2 – EBS only types general purpose: t,m compute optimized: c,h memory:r,x,u stotage:i Iam role EC2 : 169…. /meta-data Offers on demand: dev/test reserved:…
Cloud Front CloudFront Functions These two types of functions can be executed during the processing of incoming or outgoing requests in AWS, and are integrated within CloudFront. cloud functions javascript < 1 ms cache key header rewrite no access to resource like subnet, database lambda edge functions change request and response 5-10s access to…
lambda (Python, C#, Ruby, Java, rust, Go lang) execution time : 15 min max use layers Throttle error 429 512 Mo RAM max 1000 max in //
I’m starting a series of posts sharing key things to know, tips and tricks to help you succeed in the AWS Solutions Architect Associate certification. The first thing is to learn and practice. For me, it started with a course by by Stephane Marek. Then I trained regularly and took notes. I did some hands-on…
I recently took part in a hackathon at work. Here’s the idea I presented. I called my project Dentaface. The goal of this project is to automate the detection of dental cavities with X-ray images or pictures using a deep learning model (Detectron2 RetinaNet) and expose it via a FastAPI-based backend. It enables real-time caries…
I spent almost 100 hours preparing this certification, using the following Udemy courses: Test Training: Some of them are deprecated but it helps. https://www.udemy.com/course-dashboard-redirect/?course_id=2597160 https://www.udemy.com/course-dashboard-redirect/?course_id=5604244 Course and training Very useful, the first section and the pdf with tricks and tips is very useful https://www.udemy.com/course-dashboard-redirect/?course_id=6323177 Course Very long but complete. https://www.udemy.com/course-dashboard-redirect/?course_id=2196488