Denver AWS Users' Group Meeting - July 2018 Slides
0 likes80 views
This document summarizes the July 2018 meeting of the AWS Users' Group in Denver. It thanks sponsors Eplexity and Prestige for hosting the event and providing food and drinks. The agenda includes discussions on methodology changes when moving applications to the cloud as well as cloud optimization best practices. Attendees are provided WiFi login details and reminded to follow the group on Twitter. Upcoming events and resources are also listed.
Denver AWS Users' Group Meeting - July 2018 Slides
- 1. Thanks to our Sponsors, Eplexity and Prestige! Please thank them for the location, food and beverages! WiFi for members: Network: The Office @ DTC Password: office@dtc
- 2. AWS Users’ Group July 2018! David “Mac” McDaniel Sr. Solution & Cloud Architect Director of Professional Services, Eplexity david@mobile-360.com & David.McDaniel@Eplexity.com LinkedIn: https://www.linkedin.com/in/davidbmcdaniel Twitter: @ServerlessJava, @DenverAWSUG Twitter list: https://twitter.com/CloudKegGuy/lists/aws/members
- 3. Communication Items Slack Channel: https://DenverAWSUsersGroup.slack.com, email me at david@mobile-360.com for an invite. We now have our own Twitter handle: @DenverAWSUG. Please follow and we’ll post updates and surveys there as well! We are now listed on AWS UG site: https://aws.amazon.com/usergroups/americas/ Announcements: Please email me announcements at least the day before the Meetup!
- 4. Recent AWS Updates Translate Adds 6 New Languages Transcribe to identify speakers based on input channel Comprehend supports syntax analysis S3 faster SBE1 / Snowball Edge instances Bring your own IP to VPC API Gateway Private/VPC APIs AWS Landing Zone Added Cognito security JupyterHub on EMR 5.14.0 AppStream 2.0 enhancements AWS LimitMonitor w/notifications AWS AppSync - Managed GraphQL
- 5. I’ll discuss On-Prem to Cloud Methodology changes. Cloud is not just a place, but a new Methodology in how many things we’ve done for years basically the same. Mark Ferguson of Prestige Services will discuss Cloud Optimization. Optimization is another new aspect of Cloud Computing and something that can cost or save you real money. Tonight’s Topics: Methodology and Optimization
- 6. What Changes? “I thought Lift and Shift kept everything the same?” We’ll discuss: ● Operational Changes ● “Blinking Light” Changes ● Failure Points ● Backups - what are those? ● Security
- 7. Operational Changes ● Initial Server Authentication ○ Key Pairs ○ User Names ○ Passwords ○ Windows vs. Linux ● Connection Troubleshooting ○ “Firewalls” ○ “ACLs” ○ “Gateways” ● Monitoring ○ Continue current methods, but supplement with CloudWatch Logs & Events ○ Can add SIEM, APM and other types of monitoring using existing logging ○ For EC2, install agent ○ For Managed Services, configure CloudWatch/CloudTrail
- 8. “Blinking Light” Changes ● Can’t see or touch the servers, switches, etc ● They actually don’t exist, mostly - so much is virtualzied ● Most lights have actually been removed ● Have to find alternatives - CloudWatch, Healthcheck, etc. ● You will learn to trust the Console, just like a pilot trusts the guages ● Implement multiple, layered approaches to your own blinking lights
- 9. Failure Points ● Unfortunately, most failure points still exist ● For “Major” Outages, most sites quote 5-6 instances, ever ● Many smaller, regional, or service-specific issues ● Learn! Plan for failure… failure to plan will result in failure! ● But, they are much less likely to fail: ○ EBS (disk volumes) 20 times more reliable ■ 4% average AFR (annual failure rate) ■ EBS: 0.1-0.2% ○ Can’t find network failure since 2011 ○ One major S3 failure in 2017 - human error ● Power outages most likely cause ○ Transformer failures ○ Storms ● Most managed services replicate by default more than any on-prem data center would ever do. Know your service!
- 10. ?? Backups ?? ● What are they? ● Seriously, in general: ● Snapshots replace “Image” or “System” backups ● Snapshots capture entire state of machine, except for software doing other than (generally) append operations ○ Databases - SQL or otherwise ○ Anything doing block or raw I/O directly ● Keep doing database backups ● Don’t use “No Reboot” option if possible ● SO Easy to Test! That’s T-E-S-T, yes actually use, your backups!
- 11. Next in Series Part 2: What is the Cloud and Why Do I Need It? Part 3: Beyond Lift and Shift: Cloud Native
- 12. Upcoming Meetups ● August 15, 2018 - What is the Cloud and Why do I need it? ● September 19, 2018 - TBD - volunteers?
- 13. Resources ● My Blog: ○ https://davidmcdaniel.me ● AWS Blog: ○ https://aws.amazon.com/blogs/aws ● Curated Twitter List: ○ https://twitter.com/Chris_Kasten/lists/aws ● AWS Podcasts: ○ https://aws.amazon.com/podcasts/aws-podcast
- 14. ServerlessDays Denver! AWS has formally taken over organization of ServerlessDays worldwide and I’ve been asked to run the Denver ServerlessDays event! “One Day, One Track, One Community.” Details: September/October 2018, downtown Denver 150-200 attendees; 10-20 speakers; CFP to be announced as soon as date picked. One of nine North American locations; four South American; ~14 in Europe. Sponsorship opportunities available! Previously called JeffConf
- 15. Sponsorship - Eplexity Our new Primary Sponsor will be Eplexity. I am now Director of Professional Services for Eplexity. Eplexity completes the loop on software and infrastructure architect / build / manage lifecycle as Eplexity will now provide all of these services to our customers. I focus on the software architect & build aspects. We have 24x7x52 managed services including monitoring, remediation, backups etc. Fully staffed NOC with AWS trained and certified resources.
- 16. Questions?