Skip to content

Latest commit

 

History

History
78 lines (52 loc) · 4.17 KB

README.md

File metadata and controls

78 lines (52 loc) · 4.17 KB
Raw

lc-api

Limpidchart API service

gRPC API

Server implements ChartAPI service from lc-proto and Health service from grpc.health.v1.
ChartAPI service is available on 0.0.0.0:54010 and that can be configured via LC_API_GRPC_ADDRESS environment variable. Health service is available on 0.0.0.0:54011 and that can be configured via LC_API_GRPC_HEALTH_CHECK_ADDRESS environment variable.

REST API

REST API has its Swagger spec generated in ./api directory, and it's available on 0.0.0.0:54012 and that can be configured via LC_API_HTTP_ADDRESS environment variable.

REST API responses compression

You can specify the needed compression algorithm in Accept-Encoding header. Currently, only gzip and deflate are supported.
API will compress its response with the best compression using the specified algorithm.

Installation

Application needs a running instance of lc-renderer on dns:///localhost:54020 and that can be configured via LC_API_RENDERER_ADDRESS environment variable.

Both of them can be run as containers from the following images:

ghcr.io/limpidchart/lc-api:latest
ghcr.io/limpidchart/lc-renderer:latest

Latest tag always points to the latest stable release (tagged with vX.Y.Z version). If you need a dev version please use develop tag.

You can also compile and run limpidchart applications without containers. Check Dockerfile for the actual commands used for compilation.

Configuration

Application is configured via environment variables. Names of all variables and their default values are:

LC_API_RENDERER_ADDRESS=dns:///localhost:54020
LC_API_RENDERER_CONN_TIMEOUT=5
LC_API_RENDERER_REQUEST_TIMEOUT=30

LC_API_GRPC_ADDRESS=0.0.0.0:54010
LC_API_GRPC_SHUTDOWN_TIMEOUT=5

LC_API_GRPC_HEALTH_CHECK_ADDRESS=0.0.0.0:54011

LC_API_HTTP_ADDRESS=0.0.0.0:54012
LC_API_HTTP_SHUTDOWN_TIMEOUT=5
LC_API_HTTP_READ_TIMEOUT=5
LC_API_HTTP_WRITE_TIMEOUT=10
LC_API_HTTP_IDLE_TIMEOUT=120

LC_METRICS_ADDRESS=0.0.0.0:54013
LC_METRICS_SHUTDOWN_TIMEOUT=5
LC_METRICS_READ_TIMEOUT=5
LC_METRICS_WRITE_TIMEOUT=10
LC_METRICS_IDLE_TIMEOUT=120

Observability

You can scrap Prometheus /metrics endpoint on the LC_METRICS_ADDRESS (0.0.0.0:54013 by default).
Currently there is only a request_duration_seconds histogram with the default Prometheus buckets (.005, .01, .025, .05, .1, .25, .5, 1, 2.5, 5, 10).

You can use PromQL to build some useful visualisations from it (queries based on Weave Works article):

  • sum(irate(request_duration_seconds_count{job="default/hello"}[1m])) by (status_code) - instantaneous QPS of the service, by status code, aggregated across all replicas;
  • sum(irate(request_duration_seconds_count{job="default/hello", status_code=~"5.."}[1m])) - rate of requests returning 500s;
  • sum(irate(request_duration_seconds_count{job="default/hello"}[1m])) by (instance) - QPS per instance;
  • sum(rate(request_duration_seconds_sum{job="default/hello",ws="false"}[5m])) / sum(rate(request_duration_seconds_count{job="default/hello",ws="false"}[5m])) - 5-minute moving average latency of requests to the service;
  • histogram_quantile(0.99, sum(rate(request_duration_seconds_bucket{job="default/hello",ws="false"}[5m])) by (le)) - 5-min moving 99th percentile request latency;

Every request is also logged to dev/stderr, duration field contains value in seconds:

  • gRPC example: {"level":"info","time":"2021-08-15T12:41:28Z","protocol":"grpc","request_id":"26fab748-7e41-4901-882f-dbe6babb4a6f","ip":"192.168.0.1:55722","code":"OK","method":"CreateChart","duration":0.0091757}
  • HTTP example: {"level":"info","time":"2021-08-15T12:41:28Z","protocol":"http","request_id":"343177ef-078c-4396-a8ae-71a06f42de57","ip":"192.168.0.1","user_agent":"lc-acceptance-tests","referer":"","code":201,"method":"POST","path":"/v0/charts","resp_bytes_written":12863,"duration":0.0194194}