Debunking serverless myths
•
4 likes•2,307 views
In this talk we debunk common myths and misconceptions about serverless - how cold starts works, serverless is not just about saving operational cost, think about control with responsibility, and think about vendor lock-in with the reward.
Report
Share
Debunking serverless myths
- 1. Debunking Serverless Myths Yan Cui @theburningmonk
- 2. Yan Cui http://theburningmonk.com @theburningmonk Principal Engineer @ Independent Consultant
- 5. available in Austria, Switzerland, Germany, Japan, Canada, Italy and US on 30+ platforms
- 7. follow @dazneng for updates about the engineering team We’re hiring! Visit engineering.dazn.com to learn more. WE’RE HIRING!
- 8. AWS user since 2009
- 9. AWS user since 2009
- 11. What do you mean by ‘serverless’?
- 12. “Serverless”
- 14. Gojko Adzic It is serverless the same way WiFi is wireless. http://bit.ly/2yQgwwb
- 15. Serverless means… don’t pay for it if no-one uses it don’t need to worry about scaling don’t need to provision and manage servers
- 16. “Function-as-a-Service” AWS Lambda Azure Functions Google Cloud Functions Auth0 Webtask Spotinst Functions Kubeless IBM Cloud Functions
- 17. AWS Lambda
- 18. AWS Lambda API Gateway IOT SNS Kinesis CloudWatch
- 22. Simon Wardley Serverless will fundamentally change how we build business around technology and how you code.
- 23. Why serverless?
- 24. more Scalable (and scales faster!)
- 25. Cheaper (don’t pay for idle servers)
- 26. Resilience (built-in redundancy and multi-AZ)
- 27. Secure
- 28. idea production choose language + framework master language + framework ���gure out deployment configure AMI configure ELB configure autoscaling capacity planning over-provision for launch are we doing microservices? configure CI/CD
- 29. idea production choose language + framework master language + framework figure out deployment configure AMI configure ELB configure autoscaling capacity planning over-provision for launch are we doing microservices? configure CI/CD
- 30. idea production greater Velocity from idea to product
- 32. less ops responsibility on your shoulders
- 38. serverless is NOT the goal!
- 39. build products customers love to use
- 40. test ideas against the market quickly
- 41. iterate on s
- 43. own less technology, focus on creating Business Values
- 44. own less technology, focus on creating Business Values (serverless is just a good fit for this mindset)
- 45. serverless technologies are still maturing
- 47. DEBUNKING SERVERLESS MYTHS COLD STARTS COLD STARTS overplays or underplays the role of cold starts on serverless performance MYTH #1
- 48. DEBUNKING SERVERLESS MYTHS COLD STARTS COLD STARTS overplays or underplays the role of cold starts on serverless performance MYTH #1 COSTSCOSTS misunderstands thereal cost savings fromadopting serverless MYTH #2
- 49. DEBUNKING SERVERLESS MYTHS COLD STARTS COLD STARTS overplays or underplays the role of cold starts on serverless performance MYTH #1 COSTSCOSTS misunderstands thereal cost savings fromadopting serverless MYTH #2 CONTROL not realise that control of infra requires expensive skillsets to manage MYTH #3 CONTROL
- 50. DEBUNKING SERVERLESS MYTHS COLD STARTS COLD STARTS overplays or underplays the role of cold starts on serverless performance MYTH #1 COSTSCOSTS misunderstands thereal cost savings fromadopting serverless MYTH #2 CONTROL not realise that control of infra requires expensive skillsets to manage MYTH #3 CONTROL LOCK-INnot realise that lock-in risk rewards productivity, and true lock-in is dataMYTH #4 LOCK-IN
- 51. COLD STARTSCOLD STARTS overplays or underplays the role of cold starts on serverless performance MYTH #1 LOCK-IN not realise that lock-in risk rewards productivity, and true lock-in is data MYTH #4 LOCK-INCONTROL not realise that control of infra requires expensive skillsets to manage MYTH #3 CONTROLCOSTSCOSTS misunderstands the real cost savings from adopting serverless MYTH #2
- 54. “cold starts only happen to the first request”
- 55. function invocationconcurrent execution i.e. a container
- 56. function invocationconcurrent execution i.e. a container class instance method call
- 57. Lambda scales the number of concurrent executions based on traffic
- 58. existing “containers” are reused where possible
- 59. time invocation
- 74. FREQUENCY DURATION dictated by user traffic, out of your control
- 76. “Lambda is not a suitable solution for me because of cold starts”
- 77. “what is your latency requirement?”
- 78. cold starts that don’t add to user-facing latency is generally not worth worrying about
- 79. Node.js functions, no VPC, 1GB, averages ~500ms cold start with production workload
- 80. Node.js functions, no VPC, 1GB, averages ~500ms cold start with production workload (good enough for most web applications)
- 81. sporadic spikes latency existed before Lambda
- 82. GC pauses…
- 84. slow downstream, databases, etc.
- 86. cold starts is generally not an issue if you have a steady traffic pattern
- 87. time req/s
- 90. minimise the duration of cold starts so they fall within acceptable latency range
- 91. use higher memory setting if function performs CPU intensive work
- 92. use Node.js, Python or Golang
- 95. avoid VPCs unless you need to access VPC-protected resources
- 96. COSTSCOSTS misunderstands the real cost savings from adopting serverless MYTH #2 LOCK-IN not realise that lock-in risk rewards productivity, and true lock-in is data MYTH #4 LOCK-INCONTROL not realise that control of infra requires expensive skillsets to manage MYTH #3 CONTROL
- 97. “good luck when Amazon decides to raise the price of Lambda!”
- 98. AWS announced 67 price reductions in the last 5 years, and 0 price hikes
- 100. Model Training Low-latency Prediction Serving via Batching
- 102. developer but I build CRUD web APIs, why should I care if Lambda is a bad fit for machine learning and serving realtime predictions?
- 104. https://www.doc.ic.ac.uk/~rbc/papers/fse-serverless-17.pdf “This paper presents two case industrial studies of early adopters, showing how migrating an application to the Lambda deployment architecture reduced hosting costs – by between 66% and 95%…”
- 105. if you wear your shoes on your hands, should you really be surprised your feet is cold? (hint: it’s not because the shoes are not warm!)
- 112. recruitment is one of the most important things you can do as an employee it’s also a significant investment
- 116. leverage: do more with less
- 117. idea production choose language + framework master language + framework figure out deployment configure AMI configure ELB configure autoscaling capacity planning over-provision for launch are we doing microservices? configure CI/CD
- 119. Serverless might cost you just as much, or even more, but you get so much more done
- 121. 6 developers, 6 months 95% saving against EC2 15x no. of production releases per month
- 123. CONTROL not realise that control of infra requires expensive skillsets to manage MYTH #3 CONTROL LOCK-IN not realise that lock-in risk rewards productivity, and true lock-in is data MYTH #4 LOCK-IN
- 128. Controlling your own infrastructure comes with Responsibilities
- 129. to take on these responsibilities you need to have the relevant skillsets in the organization Controlling your own infrastructure comes with Responsibilities
- 130. to take on these responsibilities you need to have the relevant skillsets in the organization Controlling your own infrastructure comes with Responsibilities ENGINEERS
- 132. to take on these responsibilities you need to have the relevant skillsets in the organization Controlling your own infrastructure comes with Responsibilities ENGINEERS ADMIN
- 133. to take on these responsibilities you need to have the relevant skillsets in the organization Controlling your own infrastructure comes with Responsibilities ENGINEERS ADMIN RECRUITMENT
- 134. to take on these responsibilities you need to have the relevant skillsets in the organization Controlling your own infrastructure comes with Responsibilities ENGINEERS ADMIN RECRUITMENT MARKET
- 136. what are you paying for?
- 137. guard against the temptation to look for control for control sake
- 138. AWS Lambda, Azure Functions, Google Cloud Functions Your custom, container-based, general purpose compute platform
- 140. LOCK-IN not realise that lock-in risk rewards productivity, and true lock-in is data MYTH #4 LOCK-IN
- 141. Lock-in risk Reward
- 142. “VENDOR LOCK-IN IS THE ROOT OF ALL EVIL”
- 143. use our private cloud instead, it’s totally not a lock-in because containers!
- 148. here’s a Ferrari, have at it but I don’t know how to drive! I’ll take a taxi instead…
- 149. here’s a Ferrari, have at it but I don’t know how to drive! I’ll take a taxi instead… here’s a million lines of open source code, have at it but I don’t know how to run it! I’ll use a managed service instead…
- 150. even open-source solutions require expertise to operate, and expertise are both rare and expensive
- 151. to take on these responsibilities you need to have the relevant skills sets in the organization Controlling your own infrastructure comes with Responsibilities ENGINEERS ADMIN RECRUITMENT MARKET
- 152. if you already have the expertise, can their time be better used to generate more business value instead?
- 153. risk is just one side of the coin
- 155. extracting maximum value from your cloud provider minimising undifferentiated heavy-lifting faster time-to-market reward
- 156. vendor lock-in is a problem that is all bark but no bite
- 158. own less technology, focus on creating Business Values
- 160. “DATABASE LOCK-IN IS THE ROOT OF ALL EVIL”
- 161. ORM
- 162. a load of unnecessary complexity that didn’t end up making DB migrations any easier, and forced you to the least common denominator of feature sets and stop you from taking advantage of your DB in the first place…
- 163. nooooooooooo, wish I hadn’t used an ORM!!
- 165. The true danger with lock-in, especially with serverless, is the potential for data lock-in. Data has gravity. It accumulates. Data is economically disincentivized to leave, by way of platform pricing. This is the single biggest threat to vendor choice.
- 166. is vendor lock-in a risk? YES
- 167. is the return worth the risk? ABSOLUTELY!
- 168. there are no silver bullets
- 171. control comes with its own baggages
- 172. but sometimes the benefits outweigh the baggages