Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

How to build observability into a serverless application

707 views

Published on

Serverless introduces a number of challenges to existing tools for observability, we need to adapt our practices to fit this new paradigm. In this talk we will discuss how we can build observability into a serverless application. We will see how you can implement log aggregation, distributed tracing and correlation IDs through both synchronous as well as asynchronous events.

Published in: Technology
  • Be the first to comment

How to build observability into a serverless application

  1. 1. how to build Serverless OBSERVABILITY into a application
  2. 2. Abraham Wald
  3. 3. Abraham Wald
  4. 4. Abraham Wald
  5. 5. Abraham Wald Wald noted that the study only considered the aircraft that had survived their missions—the bombers that had been shot down were not present for the damage assessment. The holes in the returning aircraft, then, represented areas where a bomber could take damage and still return home safely.
  6. 6. Abraham Wald Wald noted that the study only considered the aircraft that had survived their missions—the bombers that had been shot down were not present for the damage assessment. The holes in the returning aircraft, then, represented areas where a bomber could take damage and still return home safely.
  7. 7. survivor bias in monitoring
  8. 8. survivor bias in monitoring Only focus on failure modes that we were able to successfully identify through investigation and postmortem in the past. The bullet holes that shot us down and we couldn’t identify stay invisible, and will continue to shoot us down.
  9. 9. What do I mean by “observability”?
  10. 10. Monitoring watching out for known failure modes in the system, e.g. network I/O, CPU, memory usage, …
  11. 11. Observability being able to debug the system, and gain insights into the system’s behaviour
  12. 12. In control theory, observability is a measure of how well internal states of a system can be inferred from knowledge of its external outputs. https://en.wikipedia.org/wiki/Observability
  13. 13. Known Success
  14. 14. Known SuccessKnown Errors
  15. 15. Known SuccessKnown Errors easy to monitor!
  16. 16. Known SuccessKnown Errors Known Unknowns
  17. 17. Known SuccessKnown Errors Known UnknownsUnknown Unknowns
  18. 18. Known SuccessKnown Errors Known UnknownsUnknown Unknowns invisible bullet holes
  19. 19. Known SuccessKnown Errors Known UnknownsUnknown Unknowns
  20. 20. Known SuccessKnown Errors Known UnknownsUnknown Unknowns alert on this
  21. 21. Known SuccessKnown Errors Known UnknownsUnknown Unknowns alert on the absence of this!
  22. 22. Known SuccessKnown Errors Known UnknownsUnknown Unknowns what went wrong?
  23. 23. These are the four pillars of the Observability Engineering team’s charter: • Monitoring • Alerting/Visualization • Distributed systems tracing infrastructure • Log aggregation/analytics “ ” http://bit.ly/2DnjyuW- Observability Engineering at Twitter
  24. 24. microservices death stars circa 2015
  25. 25. microservices death stars circa 2015 mm… I wonder what’s going on here…
  26. 26. microservices death stars circa 2015 I got this!
  27. 27. hi, my name is Yan.
  28. 28. I’m a principal engineer at
  29. 29. available in: Austria, Switzerland, Germany, Japan, Canada and Italy 30+ platforms
  30. 30. over 500,000 concurrent viewers
  31. 31. coming to the US
  32. 32. We’re hiring! Visit engineering.dazn.com to learn more. follow @dazneng for updates about the engineering team
  33. 33. follow @dazneng for updates about the engineering team We’re hiring! Visit engineering.dazn.com to learn more. WE’RE HIRING!
  34. 34. AWS user since 2009
  35. 35. http://bit.ly/yubl-serverless
  36. 36. http://bit.ly/production-ready-serverless
  37. 37. since July 2018
  38. 38. new challenges
  39. 39. NO ACCESS to underlying OS
  40. 40. NOWHERE to install agents/daemons
  41. 41. •nowhere to install agents/daemons new challenges
  42. 42. user request user request user request user request user request user request user request critical paths: minimise user-facing latency handler handler handler handler handler handler handler
  43. 43. user request user request user request user request user request user request user request critical paths: minimise user-facing latency StatsD handler handler handler handler handler handler handler rsyslog background processing: batched, asynchronous, low overhead
  44. 44. user request user request user request user request user request user request user request critical paths: minimise user-facing latency StatsD handler handler handler handler handler handler handler rsyslog background processing: batched, asynchronous, low overhead NO background processing except what platform provides
  45. 45. •no background processing •nowhere to install agents/daemons new challenges
  46. 46. EC2 concurrency used to be handled by your code
  47. 47. EC2 Lambda Lambda Lambda Lambda Lambda now, it’s handled by the AWS Lambda platform
  48. 48. EC2 logs & metrics used to be batched here
  49. 49. EC2 Lambda Lambda Lambda Lambda Lambda now, they are batched in each concurrent execution, at best…
  50. 50. HIGHER concurrency to log aggregation/telemetry system
  51. 51. •higher concurrency to telemetry system •nowhere to install agents/daemons •no background processing new challenges
  52. 52. Lambda cold start
  53. 53. Lambda data is batched between invocations
  54. 54. Lambda idle data is batched between invocations
  55. 55. Lambda idle garbage collectiondata is batched between invocations
  56. 56. Lambda idle garbage collectiondata is batched between invocations HIGH chance of data loss
  57. 57. •high chance of data loss (if batching) •nowhere to install agents/daemons •no background processing •higher concurrency to telemetry system new challenges
  58. 58. Lambda
  59. 59. my code send metrics
  60. 60. my code send metrics
  61. 61. my code send metrics internet internet press button something happens
  62. 62. http://bit.ly/2Dpidje
  63. 63. ? functions are often chained together via asynchronous invocations
  64. 64. ? SNS Kinesis CloudWatch Events CloudWatch LogsIoT DynamoDB S3 SES
  65. 65. ? SNS Kinesis CloudWatch Events CloudWatch LogsIoT DynamoDB S3 SES tracing ASYNCHRONOUS invocations through so many different event sources is difficult
  66. 66. •asynchronous invocations •nowhere to install agents/daemons •no background processing •higher concurrency to telemetry system •high chance of data loss (if batching) new challenges
  67. 67. These are the four pillars of the Observability Engineering team’s charter: • Monitoring • Alerting/Visualization • Distributed systems tracing infrastructure • Log aggregation/analytics “ ” http://bit.ly/2DnjyuW- Observability Engineering at Twitter
  68. 68. LOGGING
  69. 69. 2016-07-12T12:24:37.571Z 994f18f9-482b-11e6-8668-53e4eab441ae GOT is off air, what do I do now?
  70. 70. 2016-07-12T12:24:37.571Z 994f18f9-482b-11e6-8668-53e4eab441ae GOT is off air, what do I do now? UTC Timestamp Request Id your log message
  71. 71. one log group per function one log stream for each concurrent invocation
  72. 72. logs are not easily searchable in CloudWatch Logs me
  73. 73. CloudWatch Logs
  74. 74. CloudWatch Logs is an async event source for Lambda
  75. 75. Concurrent Executions Time regional max concurrency functions that are delivering business value
  76. 76. Concurrent Executions Time regional max concurrency functions that are delivering business value ship logs
  77. 77. either set concurrency limit on the log shipping function (and potentially lose logs due to throttling) or…
  78. 78. 1 shard = 1 concurrent execution i.e. control the no. of concurrent executions with no. of shards
  79. 79.
  80. 80. CloudWatch Logs
  81. 81. CloudWatch Logs
  82. 82. use structured logging with JSON
  83. 83. https://stackify.com/what-is-structured-logging-and-why-developers-need-it/ https://blog.treasuredata.com/blog/2012/04/26/log-everything-as-json/
  84. 84. https://www.loggly.com/blog/8-handy-tips-consider-logging-json/
  85. 85. traditional loggers are too heavy for Lambda
  86. 86. CloudWatch Logs $0.50 per GB ingested $0.03 per GB archived per month
  87. 87. CloudWatch Logs $0.50 per GB ingested $0.03 per GB archived per month 1M invocation of a 128MB function = $0.000000208 * 1M + $0.20 = $0.408
  88. 88. DON’T leave debug logging ON in production
  89. 89. have to redeploy ALL the functions along the call path to collect all relevant debug logs
  90. 90. EC2 Lambda Lambda Lambda Lambda Lambda Concurrency is handled by the AWS Lambda platform
  91. 91. sampling decision has to be followed by an entire call chain
  92. 92. Initial Request ID User ID Session ID User-Agent Order ID …
  93. 93. nonintrusive extensible consistent works for streams
  94. 94. EC2 Lambda Lambda Lambda Lambda Lambda Concurrency is handled by the AWS Lambda platform
  95. 95. store correlation IDs in global variable
  96. 96. use middleware to auto-capture incoming correlation IDs
  97. 97. extract correlation IDs from invocation event, and store them in the correlation-ids module reset
  98. 98. logger to always include captured correlation IDs
  99. 99. HTTP and AWS SDK clients to auto-forward correlation IDs on
  100. 100. context.awsRequestId get-index
  101. 101. context.awsRequestId x-correlation-id get-index
  102. 102. { “headers”: { “x-correlation-id”: “…” }, … } get-index
  103. 103. { “body”: null, “resource”: “/restaurants”, “headers”: { “x-correlation-id”: “…” }, … } get-index get-restaurants
  104. 104. get-restaurants global.CONTEXT global.CONTEXT x-correlation-id = … x-correlation-xxx = … get-index headers[“User-Agent”] headers[“Debug-Log-Enabled”] headers[“User-Agent”] headers[“Debug-Log-Enabled”] headers[“x-correlation-id”] capture forward function event log.info(…)
  105. 105. nonintrusive extensible consistent works for streams
  106. 106. MONITORING
  107. 107. •no background processing •nowhere to install agents/daemons new challenges
  108. 108. my code send metrics internet internet press button something happens
  109. 109. those extra 10-20ms for sending custom metrics would compound when you have microservices and multiple APIs are called within one slice of user event
  110. 110. Amazon found every 100ms of latency cost them 1% in sales. http://bit.ly/2EXPfbA
  111. 111. console.log(“hydrating yubls from db…”); console.log(“fetching user info from user-api”); console.log(“MONITORING|1489795335|27.4|latency|user-api-latency”); console.log(“MONITORING|1489795335|8|count|yubls-served”); timestamp metric value metric type metric namemetrics logs
  112. 112. CloudWatch Logs AWS Lambda ELK stack logs m etrics CloudWatch
  113. 113. delay cost concurrency
  114. 114. delay cost concurrency no latency overhead
  115. 115. API Gateway send custom metrics asynchronously
  116. 116. SNS KinesisS3API Gateway … send custom metrics asynchronously send custom metrics as part of function invocation
  117. 117. TRACING
  118. 118. X-Ray
  119. 119. don’t span over async invocations good for identifying dependencies of a function, but not good enough for tracing the entire call chain as user request/data flows through the system via async event sources.
  120. 120. don’t span over non-AWS services
  121. 121. write structured logs
  122. 122. instrument your code
  123. 123. make it easy to do the right thing
  124. 124. Yan Cui http://theburningmonk.com @theburningmonk
  125. 125. follow @dazneng for updates about the engineering team We’re hiring! Visit engineering.dazn.com to learn more. WE’RE HIRING!

×