AzurePublicDataset/AzureFunctionsInvocationTra...

3.3 KiB

Azure Functions Invocation Trace 2021

-- revision 1, 2021-11-30

Introduction

This is a trace of function invocations in Microsoft's Azure Functions for two weeks starting on 2021-01-31. This trace has been used in the SOSP 2021 paper Faster and Cheaper Serverless Computing on Harvested Resources.

Using the Data

License

The data is made available and licensed under a CC-BY Attribution License. By downloading it or using them, you agree to the terms of this license.

Attribution

If you use this data for a publication or project, please cite the accompanying paper:

Yanqi Zhang, Íñigo Goiri, Gohar Irfan Chaudhry, Rodrigo Fonseca, Sameh Elnikety, Christina Delimitrou, Ricardo Bianchini. "Faster and Cheaper Serverless Computing on Harvested Resources", in Proceedings of the ACM International Symposium on Operating Systems Principles (SOSP), October 2021.

If you have any questions, comments, or concerns, or if you would like to share tools for working with the traces, please contact us at azurepublicdataset@service.microsoft.com

Downloading

You can download the dataset here: AzureFunctionsInvocationTraceForTwoWeeksJan2021.rar.

Schema and Description

Schema

  • app: application id (encrypted)
  • func: function id (encrypted), and unique only within an application
  • end_timestamp: function invocation end timestamp (in seconds)
  • duration: duration of function invocation (in seconds)

Remarks

In Azure Functions, the unit of deployment is called an application, and an application has one or more functions. For example, an application could be a binary file with one or more entry points. A function invocation specifies both the app id and the func id withen the app.

Invocation timestamps have been modified from those in the actual production trace.

Sample

app func end_timestamp duration
734272c01926d19690e5ec308bab64ef97950b75b1c7582283e0783fce1751d8 313c03f53a0d31f70aec25f62efb33e7dd779725ca4af579018452d1204beaad 5160.142570018768 0.134
17c37a0fdd5d1932b755c0e6447137bc08fd524f455e14fdac414f584de08dc5 c9f8e30e36d1aef62c10b3cfca6e289a93848a148d876dd514753040314f4817 5161.280997037888 0.013
7fa05b607ae861b85ec53cea12d3efaed8be0f9a92f5d6e8067244161d491e96 9bc86d6cd1ee254aaa313492f0fd88be8bd7b92d50d4237ff52d7685440c0906 5241.567729949951 42.356
c8c43e1a911f29e5506460a2fbef61ff39723d672f3b3b67d12d4c236c6872f7 653cdbc309bc359f3289d3b4df21c4a8e478d22946b35cbfdab05377dcacd3e0 5253.883348941803 42.372
db6be4a997f386b37c6246aaeecf81ab81562db84cf4c0d44907d9df2d0ab9fc 9040b71f8a0325ba418c85bcefa3b19c02c781bed6284af487d3f111f369534a 5219.518173933029 0.108
f7bfe5bc8d2a37a5c15986fbfc2c477a746e866adcb9663f9df7535b61c3eb9b 34f4775366e51728635af48df1a96d332cf1565eee069a0030f12966ae760274 5220.1072909832 0.093