Today I attended Microsoft Ignite in Sydney. It's a 2 day event aimed at showcasing services offered by Microsoft and how they can be leveraged to solve various business problems.
Following is a braindump of my key takeaways:
Session 1: 10 Best security practices in Azure
Speaker: Abbas Kudrati
1. Azure Security center: Use the security score provided to view the current security status
2. Admin account protection
Multi factor authentication
No standing access
3. Enterprise segmentation and Zero trust preparation
Monitor for attacks
Push logs to azure security Center
4. Centralized visibility
Azure security and azure sentinel
5. Secure Devops
6. Security at the Core GRC - key responsible parties
7. Network and Containment
Choose firewall strategy
8. Use web app firewall on all internet facing applications
9. DDoS mitigation
Azure services comes pre-built with DDoS mitigation
10. Network
Deprecating legacy technology
Session 2: 10 amazing things about Edge
Speaker: Vyankatesh Gadekar
10. Enterprise new tab page
9. Microsoft search on Bing: find your documents within the company in Bing
8. Best in class developer tools - chromium as a baseline. 7. Tracking prevention - Privacy dashboard is built into the browser. 6. Multiple profiles and AAD sign in with SSO
5. Enterprise sync
4. Security stack:
SmartScreen
Application Guard - run a untrusted site in a container that is isolated from the local device conditional access - accessing a resource will depend on company policies
3. Integrated deployment with ConfigMgr or Intune
2. App assure: If your web apps works in Chrome Or IE 11, then there is a assurance that it will work in Edge
1. IE mode: is integrated into Microsoft edge. It will open all legacy websites.
Session 3: Deployment and A/B testing
Speaker: Michael Ridland
Presentation Slides
Various deployment slots helps to deploy to an inactive slot first, do the necessary testing and then swap those slots.
Canary testing: You can maintain multiple production slots and define the traffic flow for each
Why?
Validate changes before production
Eliminate downtime
Roll back when necessary
You can deploy into a slot as part of your build pipeline.
Session 4: Using pre-built AI to solve business problems
Speaker: David Smith
Keywords
Azure cognitive services
Custom Vision service - this is used to train the model
Trained convolutional NN
Training an image classifier
Transfer learning
Microsoft Personalizer: It is not a recommendation engine, it uses the neural network. It has exploit and explore mode. The idea is to exploit for performance and explore for alternate options
When working with AI, consider the following:
Enable people
Be inclusive
Be fair and transparent
Session 5: Containerization
Speaker: Rohan Arora
Containers helps to set up clear boundaries for your applications
Why Containers?
No more “works on my machine”
Every environment looks the same
Increased velocity
Docker: It is most popular container runtime. It allows to build applications, test them in local docker environment on the computer and deploy the same image to the cloud.
Dockerfile: Instructions to compose your environment
Here are some happy snaps from the day at the Telstra Purple booth :)
Comments