Episode 82: Attack of the two-pizza teams
Software Defined Talk
English - December 08, 2016 23:00 - 57 minutes - 27.1 MB - ★★★★★ - 64 ratingsTech News News Technology Homepage Download Apple Podcasts Google Podcasts Overcast Castro Pocket Casts RSS feed
...Eventually, someone has to clean up the leftover pizza.
...That sweet OpEx.
..."Easy to stay."
Amazon came out with a slew of features last week. This week we discuss them and take some cracks at the broad, portfolio approach at AWS compared to historic (like .Net) platform approaches. We also discuss footwear and what to eat and where to stay in Las Vegas.
Footware
Kenneth Cole slip on shoes (http://amzn.to/2gH6OzD).
Keen Austin shoes, slip-on (http://amzn.to/2h2gveX) and lace (http://amzn.to/2ggll4y).
The Doc Martin's Coté used to wear, Hickmire (http://amzn.to/2hlPnIJ).
Mid-roll
Coté: the Cloud Native roadshows are over, but check out the cloud native WIP I have at cote.io/cloud2 (http://cote.io/cloud2) or, just check out some excerpts on working with auditors (https://medium.com/@cote/auditors-your-new-bffs-918c8671897a#.et5tv7p7l), selecting initial projects (https://medium.com/@cote/getting-started-picking-your-first-cloud-native-projects-or-every-digital-transformation-starts-d0b1295f3712#.v7jpyjvro), and dealing with legacy (https://medium.com/built-to-adapt/deal-with-legacy-before-it-deals-with-you-cc907c800845#.ixtz1kqdz).
Matt: Presenting at the CC Dojo #3, talking DevOps in Tokyo (https://connpass.com/event/46308/)
AWS re:Invent
Matt Ray heroically summarizes all here.
Richard has a write-up as well (https://www.infoq.com/news/2016/12/aws-reinvent-recap).
RedMonk re:Cap (http://redmonk.com/sogrady/2016/12/07/the-redmonk-reinvent-recap/)
Global Partner Summit
Don't hedge your bets, "AWS has no time for uncommitted partners" (http://www.zdnet.com/article/andy-jassy-warns-aws-has-no-time-for-uncommitted-partners/)
"10,000 new Partners have joined the APN in the past 12 months" (https://aws.amazon.com/blogs/aws/aws-global-partner-summit-report-from-reinvent-2016/)
Day 1 - "I'd like to tell you about…"
Amazon Lightsail (https://aws.amazon.com/blogs/aws/amazon-lightsail-the-power-of-aws-the-simplicity-of-a-vps/)
Monthly instances with memory, cpu, storage & static IP
Bitnami! Hello Digital Ocean & Linode
Amazon Athena (https://aws.amazon.com/blogs/aws/amazon-athena-interactive-sql-queries-for-data-in-amazon-s3/)
S3 SQL queries, based on Presto distributed SQL engine
JSON, CSV, log files, delimited text, others
Coté: this seems pretty amazing.
Amazon Rekognition (https://aws.amazon.com/blogs/aws/amazon-rekognition-image-detection-and-recognition-powered-by-deep-learning/)
Image detection & recognition
Amazon Polly (https://aws.amazon.com/blogs/aws/polly-text-to-speech-in-47-voices-and-24-languages/)
Text to Speech in 47 Voices and 24 Languages
Coté: Makes transcripts?
Amazon Lex (https://aws.amazon.com/blogs/aws/amazon-lex-build-conversational-voice-text-interfaces/)
Conversational voice & text interface builder (ie. chatbots)
Coté: make chat-bots and such.
AWS Greengrass (https://aws.amazon.com/blogs/aws/aws-greengrass-ubiquitous-real-world-computing/)
Local Lambda processing for IoT
Coté: is this supposed to be, like, for running Lambda things on disconnected devices? Like fPaaS in my car?
AWS Snowball Edge & Snowmobile (https://aws.amazon.com/blogs/aws/aws-snowball-edge-more-storage-local-endpoints-lambda-functions/)
Local processing of data? S3/NFS and local Lambda processing? I'm thinking easy hybrid on-ramp
Not just me (https://twitter.com/CTOAdvisor/status/806320423881162753)
More on it (http://www.techrepublic.com/article/how-amazon-is-moving-closer-to-on-premises-compute-with-snowball-edge/)
Move exabytes in weeks (https://aws.amazon.com/blogs/aws/aws-snowmobile-move-exabytes-of-data-to-the-cloud-in-weeks/)
"Snowmobile is a ruggedized, tamper-resistant shipping container 45 feet long, 9.6 feet high, and 8 feet wide. It is waterproof, climate-controlled, and can be parked in a covered or uncovered area adjacent to your existing data center."
Coté: LEGOS!
More instance types, Elastic GPUs, F1 Instances, PostgreSQL for Aurora
High I/O (I3 3.3 million IOPs 16GB/s), compute (C5 72 vCPUs, 144 GiB), memory (R4 488 Gib), burstable (T2 shared) (https://aws.amazon.com/blogs/aws/ec2-instance-type-update-t2-r4-f1-elastic-gpus-i3-c5/)
Mix EC2 instance type with a 1-8 GiB GPU (https://aws.amazon.com/blogs/aws/in-the-work-amazon-ec2-elastic-gpus/)
More! (https://aws.amazon.com/blogs/aws/developer-preview-ec2-instances-f1-with-programmable-hardware/)
F1: FPGA EC2 instances, also available for use in the AWS Marketplace (https://aws.amazon.com/blogs/aws/amazon-aurora-update-postgresql-compatibility/)
RDS vs. Aurora Postgres? Aurora is more fault tolerant apparently?
Day 2
AWS OpsWorks for Chef Automate (https://aws.amazon.com/opsworks/chefautomate/)
Chef blog (https://blog.chef.io/2016/12/01/chef-automate-now-available-fully-managed-service-aws/)
Fully managed Chef Server & Automate
Previous OpsWorks now called "OpsWorks Stacks"
Cloud Opinion approves the Chef strategy (https://twitter.com/cloud_opinion/status/804374597449584640)
EC2 Systems Manager
Tools for managing EC2 & on-premises systems (https://aws.amazon.com/ec2/systems-manager/)
AWS Codebuild
Managed elastic build service with testing (https://aws.amazon.com/blogs/aws/aws-codebuild-fully-managed-build-service/)
AWS X-Ray (https://aws.amazon.com/blogs/aws/aws-x-ray-see-inside-of-your-distributed-application/)
Distributed debugging service for EC2/ECS/Lambda?
"easy way for developers to "follow-the-thread" as execution traverses EC2 instances, ECS containers, microservices, AWS database and messaging services"
AWS Personal Health Dashboard (https://aws.amazon.com/blogs/aws/new-aws-personal-health-dashboard-status-you-can-relate-to/)
Personalized AWS monitoring & CloudWatch Events auto-remediation
Disruptive to PAAS monitoring & APM (New Relic, DataDog, App Dynamics)
AWS Shield (https://aws.amazon.com/blogs/aws/aws-shield-protect-your-applications-from-ddos-attacks/)
DDoS protection
Amazon Pinpoint
Mobile notification & analytics service (https://aws.amazon.com/blogs/aws/amazon-pinpoint-hit-your-targets-with-aws/)
AWS Glue
Managed data catalog & ETL (extract, transform & load) service for data analysis
AWS Batch
Automated AWS provisioning for batch jobs (https://aws.amazon.com/blogs/aws/aws-batch-run-batch-computing-jobs-on-aws/)
C# in Lamba, Lambda Edge, AWS Step Functions
Werner Vogels: "serverless, there is no cattle, only the herd"
Lambda Edge (https://aws.amazon.com/blogs/aws/coming-soon-lambda-at-the-edge/) for running in response to CloudFront events, ""intelligent" processing of HTTP requests at a location that is close"
More (https://aws.amazon.com/blogs/aws/new-aws-step-functions-build-distributed-applications-using-visual-workflows/)
Step Functions a visual workflow "state machine" for Lambda functions
More (https://serverless.zone/faas-is-stateless-and-aws-step-functions-provides-state-as-a-service-2499d4a6e412)
BLOX (https://aws.amazon.com/blogs/compute/introducing-blox-from-amazon-ec2-container-service/): EC2 Container Service Scheduler
Open source scheduler, watches CloudWatch events for managing ECS deployments
Blox.github.io
Analysis discussion for all the AWS stuff
Jesus! I couldn't read it all!
So, what's the role of Lambda here? It seems like the universal process thingy - like AppleScript, bash scripts, etc. for each part: if you need/want to add some customization to each thing, put a Lambda on it.
What's the argument against just going full Amazon, in the same way you'd go full .Net, etc.? Is it cost? Lockin? Performance (people always talk about Amazon being kind of flakey at times - but what isn't flakey, your in-house run IT? Come on.)
BONUS LINKS! Not covered in episode.
Docker for AWS
"EC2 Container Service, Elastic Beanstalk, and Docker for AWS all cost nothing; the only costs are those incurred by using AWS resources like EC2 or EBS." (http://www.infoworld.com/article/3145696/application-development/docker-for-aws-whos-it-really-for.html)
Docker gets paid on usage?
Apparently an easier learning curve than ECS + AWS services, but whither Blox?
Time to Break up Amazon?
Someone has an opinion (http://www.geekwire.com/2016/new-study-compares-amazon-19th-century-robber-barons-urges-policymakers-break-online-retail-giant/)
HPE Discover, all about the "Hybrid Cloud"
Hybrid it up! (http://www.zdnet.com/article/hpe-updates-its-converged-infrastructure-hybrid-cloud-software-lineup/)
Killed "The Machine" (http://www.theregister.co.uk/2016/11/29/hp_labs_delivered_machine_proof_of_concept_prototype_but_machine_product_is_no_more/)
HPE's Synergy software, based on OpenStack (is this just Helion rebranded?)
Not great timing for a conference
Sold OpenStack & CloudFoundry bits to SUSE (http://thenewstack.io/suse-add-hpes-openstack-cloud-foundry-portfolio-boost-kubernetes-investment/), the new "preferred Linux partner":
How Google is Challenging AWS
Ben on public cloud (https://stratechery.com/2016/how-google-cloud-platform-is-challenging-aws/)
"open-sourcing Kubernetes was Google's attempt to effectively build a browser on top of cloud infrastructure and thus decrease switching costs; the company's equivalent of Google Search will be machine learning."
Exponent.fm episode 097 — Google vs AWS (http://exponent.fm/episode-097-google-versus-aws/)
Recommendations
Brandon:
Apple Wifi Calling (https://support.apple.com/en-us/HT203032) & Airplane mode (https://support.apple.com/en-us/HT204234).
Westworld worth watching (http://www.hbo.com/westworld).
Matt:
Backyard Kookaburras (https://www.youtube.com/watch?v=DmNn7P59HcQ).
Magpies too! (http://www.musicalsoupeaters.com/swooping-season/)
This gif (https://media.giphy.com/media/wik7sKOl86OFq/giphy.gif).
Coté: W Hotel in Las Vegas (http://www.wlasvegas.com/) and lobster eggs benedict (https://www.instagram.com/p/BNxAyQbjKCQ/) at Payard's in Ceasers'
Outro: "I need my minutes," Soul Position (http://genius.com/Soul-position-i-need-my-minutes-lyrics).
...Eventually, someone has to clean up the leftover pizza.
...That sweet OpEx.
..."Easy to stay."
Amazon came out with a slew of features last week. This week we discuss them and take some cracks at the broad, portfolio approach at AWS compared to historic (like .Net) platform approaches. We also discuss footwear and what to eat and where to stay in Las Vegas.
Footware
Kenneth Cole slip on shoes.
Keen Austin shoes, slip-on and lace.
The Doc Martin's Coté used to wear, Hickmire.
Mid-roll
Coté: the Cloud Native roadshows are over, but check out the cloud native WIP I have at cote.io/cloud2 or, just check out some excerpts on working with auditors, selecting initial projects, and dealing with legacy.
Matt: Presenting at the CC Dojo #3, talking DevOps in Tokyo
AWS re:Invent
Matt Ray heroically summarizes all here.
Richard has a write-up as well.
RedMonk re:Cap
Global Partner Summit
Don't hedge your bets, "AWS has no time for uncommitted partners"
"10,000 new Partners have joined the APN in the past 12 months"
Day 1 - "I'd like to tell you about…"
Monthly instances with memory, cpu, storage & static IP
Bitnami! Hello Digital Ocean & Linode
S3 SQL queries, based on Presto distributed SQL engine
JSON, CSV, log files, delimited text, others
Coté: this seems pretty amazing.
Image detection & recognition
Text to Speech in 47 Voices and 24 Languages
Coté: Makes transcripts?
Conversational voice & text interface builder (ie. chatbots)
Coté: make chat-bots and such.
Local Lambda processing for IoT
Coté: is this supposed to be, like, for running Lambda things on disconnected devices? Like fPaaS in my car?
AWS Snowball Edge & Snowmobile
Local processing of data? S3/NFS and local Lambda processing? I'm thinking easy hybrid on-ramp
Move exabytes in weeks
"Snowmobile is a ruggedized, tamper-resistant shipping container 45 feet long, 9.6 feet high, and 8 feet wide. It is waterproof, climate-controlled, and can be parked in a covered or uncovered area adjacent to your existing data center."
Coté: LEGOS!
More instance types, Elastic GPUs, F1 Instances, PostgreSQL for Aurora
High I/O (I3 3.3 million IOPs 16GB/s), compute (C5 72 vCPUs, 144 GiB), memory (R4 488 Gib), burstable (T2 shared)
Mix EC2 instance type with a 1-8 GiB GPU
More!
F1: FPGA EC2 instances, also available for use in the AWS Marketplace
RDS vs. Aurora Postgres? Aurora is more fault tolerant apparently?
Day 2
AWS OpsWorks for Chef Automate
Chef blog
Fully managed Chef Server & Automate
Previous OpsWorks now called "OpsWorks Stacks"
Cloud Opinion approves the Chef strategy
EC2 Systems Manager
Tools for managing EC2 & on-premises systems
AWS Codebuild
Managed elastic build service with testing
Distributed debugging service for EC2/ECS/Lambda?
"easy way for developers to "follow-the-thread" as execution traverses EC2 instances, ECS containers, microservices, AWS database and messaging services"
Personalized AWS monitoring & CloudWatch Events auto-remediation
Disruptive to PAAS monitoring & APM (New Relic, DataDog, App Dynamics)
DDoS protection
Amazon Pinpoint
Mobile notification & analytics service
AWS Glue
Managed data catalog & ETL (extract, transform & load) service for data analysis
AWS Batch
Automated AWS provisioning for batch jobs
C# in Lamba, Lambda Edge, AWS Step Functions
Werner Vogels: "serverless, there is no cattle, only the herd"
Lambda Edge for running in response to CloudFront events, ""intelligent" processing of HTTP requests at a location that is close"
More
Step Functions a visual workflow "state machine" for Lambda functions
More
BLOX: EC2 Container Service Scheduler
Open source scheduler, watches CloudWatch events for managing ECS deployments
Blox.github.io
Analysis discussion for all the AWS stuff
Jesus! I couldn't read it all!
So, what's the role of Lambda here? It seems like the universal process thingy - like AppleScript, bash scripts, etc. for each part: if you need/want to add some customization to each thing, put a Lambda on it.
What's the argument against just going full Amazon, in the same way you'd go full .Net, etc.? Is it cost? Lockin? Performance (people always talk about Amazon being kind of flakey at times - but what isn't flakey, your in-house run IT? Come on.)
BONUS LINKS! Not covered in episode.
Docker for AWS
"EC2 Container Service, Elastic Beanstalk, and Docker for AWS all cost nothing; the only costs are those incurred by using AWS resources like EC2 or EBS."
Docker gets paid on usage?
Apparently an easier learning curve than ECS + AWS services, but whither Blox?
Time to Break up Amazon?
HPE Discover, all about the "Hybrid Cloud"
Hybrid it up!
Killed "The Machine"
HPE's Synergy software, based on OpenStack (is this just Helion rebranded?)
Not great timing for a conference
Sold OpenStack & CloudFoundry bits to SUSE, the new "preferred Linux partner":
How Google is Challenging AWS
Ben on public cloud
"open-sourcing Kubernetes was Google's attempt to effectively build a browser on top of cloud infrastructure and thus decrease switching costs; the company's equivalent of Google Search will be machine learning."
Exponent.fm episode 097 — Google vs AWS
Recommendations
Brandon:
Apple Wifi Calling & Airplane mode.
Westworld worth watching.
Matt:
Backyard Kookaburras.
Magpies too!
This gif.
Coté: W Hotel in Las Vegas and lobster eggs benedict at Payard's in Ceasers'
Outro: "I need my minutes," Soul Position.
Sponsored By:
Pivotal: Check out Coté's work in progress, the ~50 page cloud native journey, edition two book. It coverers the common questions, best practices, and snarky takes on doing better software in large organizations.