vision

vision-api

About

This example shows how to use both the trace and stats/exporters to instrument your backend. The API sends information both to Prometheus and to StackDriver monitoring sending spans of methods invoked.

The API server presented is one a company might use for their app and its purpose is to give them information about their backend's usage e.g:

• images processed
• methods used i.e. upload with body vs by URI
• bytes processsed altogether
• the performance and time spent within each service
• error rates

The information collected will help their engineering teams figure out what services to optimize for, what metrics their business development team will present, the popularity of their service, general performance, error rates.

Installing it

$ go get -u -v github.com/orijtech/opencensus-demos/vision

Running it

$ GOOGLE_APPLICATION_CREDENTIALS=~/Downloads/census-demos-237a8e1e41df.json vision --gcp-id census-demos

where

• GOOGLE_APPLICATION_CREDENTIALS=~/Downloads/census-demos-237a8e1e41df.json an alternate way of setting the Google authentication credentials if you didn't set them in your environment already
• --gcp-id: the ID of the project

Browser app

You can visit http://localhost:8899/

Commandline app

Please see the routes sub-section and sample responses

Routes

Sample response using JSON based route "/url"

$ curl -X POST http://localhost:8899/url --data '{"url": "https://avatars3.githubusercontent.com/u/4898263?s=460&v=4"}'

which gives

{"faces":[{"bounding_poly":{"vertices":[{"x":111,"y":70},{"x":291,"y":70},{"x":291,"y":279},{"x":111,"y":279}]},"fd_bounding_poly":{"vertices":[{"x":125,"y":122},{"x":254,"y":122},{"x":254,"y":251},{"x":125,"y":251}]},"landmarks":[{"type":1,"position":{"x":170.74551,"y":163.39682,"z":-0.00009324717}},{"type":2,"position":{"x":226.9261,"y":162.27528,"z":-3.3436992}},{"type":3,"position":{"x":153.67024,"y":150.73282,"z":6.6217027}},{"type":4,"position":{"x":184.95543,"y":150.01039,"z":-11.03244}},{"type":5,"position":{"x":211.29453,"y":149.94893,"z":-12.643309}},{"type":6,"position":{"x":244.78865,"y":151.11148,"z":1.3010101}},{"type":7,"position":{"x":197.63452,"y":162.08948,"z":-13.04061}},{"type":8,"position":{"x":197.02707,"y":190.10608,"z":-29.71414}},{"type":9,"position":{"x":196.98128,"y":211.07048,"z":-18.037405}},{"type":10,"position":{"x":197.19337,"y":230.66444,"z":-14.682273}},{"type":11,"position":{"x":172.81384,"y":220.45654,"z":-1.3366407}},{"type":12,"position":{"x":225.52855,"y":219.70993,"z":-3.4198322}},{"type":13,"position":{"x":197.22342,"y":220.48212,"z":-14.274697}},{"type":14,"position":{"x":214.58203,"y":196.89085,"z":-9.772623}},{"type":15,"position":{"x":180.6429,"y":196.00703,"z":-7.762398}},{"type":16,"position":{"x":196.71512,"y":200.28355,"z":-17.826805}},{"type":17,"position":{"x":171.36452,"y":159.27315,"z":-3.6517794}},{"type":18,"position":{"x":181.57576,"y":164.14264,"z":-0.5285912}},{"type":19,"position":{"x":170.20926,"y":166.77805,"z":-0.6425587}},{"type":20,"position":{"x":160.3396,"y":163.68492,"z":5.7165737}},{"type":29,"position":{"x":170.58768,"y":163.04672,"z":-1.5383104}},{"type":21,"position":{"x":225.21132,"y":159.61885,"z":-6.943501}},{"type":22,"position":{"x":236.61172,"y":163.82162,"z":0.9396738}},{"type":23,"position":{"x":226.92223,"y":166.23566,"z":-3.9666982}},{"type":24,"position":{"x":214.53113,"y":163.99612,"z":-2.5623853}},{"type":30,"position":{"x":225.85002,"y":163.37703,"z":-5.023317}},{"type":25,"position":{"x":169.09799,"y":143.48747,"z":-5.7417264}},{"type":26,"position":{"x":227.47415,"y":143.87534,"z":-9.315583}},{"type":27,"position":{"x":140.72449,"y":190.75964,"z":68.05329}},{"type":28,"position":{"x":263.5416,"y":192.86166,"z":60.262157}},{"type":31,"position":{"x":197.83046,"y":149.26822,"z":-13.928979}},{"type":32,"position":{"x":197.46735,"y":254.6251,"z":-8.149228}},{"type":33,"position":{"x":145.47163,"y":225.18817,"z":44.309467}},{"type":34,"position":{"x":255.8797,"y":225.8844,"z":37.550243}}],"roll_angle":0.5147558,"pan_angle":-3.5062222,"tilt_angle":2.0259361,"detection_confidence":0.8725568,"landmarking_confidence":0.5131235,"joy_likelihood":4,"sorrow_likelihood":1,"anger_likelihood":1,"surprise_likelihood":1,"under_exposed_likelihood":1,"blurred_likelihood":1,"headwear_likelihood":1}],"labels":[{"mid":"/m/04yx4","description":"man","score":0.9350149},{"mid":"/m/01g317","description":"person","score":0.93318623},{"mid":"/m/05zppz","description":"male","score":0.8540857},{"mid":"/m/08t9c_","description":"grass","score":0.73753375},{"mid":"/m/07j7r","description":"tree","score":0.7349218},{"mid":"/m/06bm2","description":"recreation","score":0.643857},{"mid":"/m/02vzx9","description":"player","score":0.63084155},{"mid":"/m/01qkbx","description":"professional","score":0.5828397},{"mid":"/m/0ds99lh","description":"fun","score":0.55307543},{"mid":"/m/05s2s","description":"plant","score":0.5302758}]}

Sample Prometheus output

# HELP opencensus_bytes_bucket_cum number of bytes ingested over time
# TYPE opencensus_bytes_bucket_cum histogram
opencensus_bytes_bucket_cum_bucket{le="0"} 0
opencensus_bytes_bucket_cum_bucket{le="1024"} 0
opencensus_bytes_bucket_cum_bucket{le="102400"} 1
opencensus_bytes_bucket_cum_bucket{le="1.048576e+06"} 1
opencensus_bytes_bucket_cum_bucket{le="1.048576e+07"} 0
opencensus_bytes_bucket_cum_bucket{le="1.048576e+08"} 0
opencensus_bytes_bucket_cum_bucket{le="1.073741824e+10"} 0
opencensus_bytes_bucket_cum_bucket{le="+Inf"} 2
opencensus_bytes_bucket_cum_sum 151606
opencensus_bytes_bucket_cum_count 2
# HELP opencensus_images_cum number of images uploaded over time
# TYPE opencensus_images_cum counter
opencensus_images_cum 2