Running Multiple Microgateway configurations on the same machine

[michaelmcdowell]

We would like to run two microgateway services on the same machine.

# 1 a token endpoint

# 2 an oauth collection of protected services

If possible, how should the <org>-<dev>-config.yaml configurations be set up?

Thanks,

Michael

[Former Community Member · Accepted solution]

See here: http://docs.apigee.com/microgateway/latest/operation-and-configuration-reference-edge-microgateway

For example: MG 1 would have

edgemicro:
  port: 8000
  max_connections: 1000
  max_connections_hard: 5000
  max_times: 300
  config_change_poll_interval: 600
  proxies:
    - edgemicro_auth 
  logging:
    level: error
    dir: /var/tmp
    stats_log_interval: 60
    rotate_interval: 24
  plugins:
    sequence:
headers:
  x-forwarded-for: true
  x-forwarded-host: true

The second MG can have:

edgemicro:
  port: 8001
  max_connections: 1000
  max_connections_hard: 5000
  max_times: 300
  config_change_poll_interval: 600
  proxies:
    - edgemicro_proxy1
    - edgemicro_proxy2
    - edgemicro_proxy3 
  logging:
    level: error
    dir: /var/tmp
    stats_log_interval: 60
    rotate_interval: 24
  plugins:
    sequence:
      - oauth
headers:
  x-forwarded-for: true
  x-forwarded-host: true

Note the differences:

1) MG does not load the oauth plugin

2) MG 1 and 2 run on different ports.

[michaelmcdowell]

... and then how might they distinguish the microgateway aware proxy (edgemicro_<unique name>) to use, specify their appropriate target url?

[Former Community Member · Accepted solution]

See here: http://docs.apigee.com/microgateway/latest/operation-and-configuration-reference-edge-microgateway

For example: MG 1 would have

edgemicro:
  port: 8000
  max_connections: 1000
  max_connections_hard: 5000
  max_times: 300
  config_change_poll_interval: 600
  proxies:
    - edgemicro_auth 
  logging:
    level: error
    dir: /var/tmp
    stats_log_interval: 60
    rotate_interval: 24
  plugins:
    sequence:
headers:
  x-forwarded-for: true
  x-forwarded-host: true

The second MG can have:

edgemicro:
  port: 8001
  max_connections: 1000
  max_connections_hard: 5000
  max_times: 300
  config_change_poll_interval: 600
  proxies:
    - edgemicro_proxy1
    - edgemicro_proxy2
    - edgemicro_proxy3 
  logging:
    level: error
    dir: /var/tmp
    stats_log_interval: 60
    rotate_interval: 24
  plugins:
    sequence:
      - oauth
headers:
  x-forwarded-for: true
  x-forwarded-host: true

Note the differences:

1) MG does not load the oauth plugin

2) MG 1 and 2 run on different ports.

[michaelmcdowell]

Are you saying I can put both these in the same <org>-dev-config.yaml file?

[Former Community Member]

No. You will need two {org}-{env}-config.yaml files. However, since it is for the same ORG/ENV, there can only be one file in a folder.

You will see the MG supports an environment variable called: EDGEMICRO_CONFIG_DIR, this variable specifies the path to the config file (default is: ~/.edgemicro).

Before starting MG1, set the variable to "/app/config1" and before starting MG2, set the variable to "/app/config2".

You don't have to use environment variables. The option is also available as a flag on start.

Ex:

edgemicro start -o org -e test -k key -s secret -c /app/config1

edgemicro start -o org -e test -k key -s secret -c /app/config2

[michaelmcdowell]

Ah, that's perfect. One final question:

How might they distinguish the Edge microgateway aware proxy (edgemicro_<unique name>) to use, specify their appropriate target url?

E.g.

• edgemicro_mg1 with HTTPTargetConnection.Url = https://foogoobar.com/v1/
• edgemicro_mg2 with HTTPTargetConnection.Url = https://baz.com/v1/

Michael McD.

[Former Community Member]

I'm afraid, I don't quite understand that question. Are you asking how to override a target endpoint or how a consumer will know which MG instance to hit?

[michaelmcdowell]

Given two microgateway aware proxies, both with unique target urls

How does a microgateway instance choose which microgateway aware proxy's URLs to use as a target?

[Former Community Member · Accepted solution]

Microgateway get target endpoint information from the API Proxy definition.

For ex:

API Proxy 1: name: edgemicro_proxy1, basePath: /api1, target: https://foo1.com

API Proxy 2: name: edgemicro)proxy2, basePath: /api2, target:https://foo2.com

When MG is accessed via localhost:8000/api1, the call is sent to foo1.com.