HTTP Server
Getting Started
This document describes the implementation of the Server version of On-Prem. See the On-Prem Overview for a higher-level introduction to On-Prem. Before provisioning servers, see the Hardware Guidance in the On-Prem Overview.
Required Software
The On-Prem HTTP Server is developed and tested against the following system environment:
- Oracle JDK 7
- Ubuntu 14.04
The following packages are also required:
- upstart (installed by default on Ubuntu)
- cronolog
Use only the Oracle JDK — issues with the SSL implementation of OpenJDK have been encountered in production deployments which can prevent On-Prem from updating its audience and geofence data.
Installation
Create a new user
First create a new user:
$ createuser factual
Download On-Prem Server
Next, download the latest On-Prem Server binary here. Unpack and copy the provided files to /usr/local/factual/outpost
.
$ mkdir -p /usr/local/factual/outpost
$ cd /usr/local/factual/outpost
$ wget [download link]
$ tar -xvzf outpost-3.1.4-example.tar.gz
$ cp -a outpost-3.1.4-example/. /usr/local/factual/outpost/
You’ll also want to create the following directory for storing logs:
$ mkdir -p /var/log/factual/outpost
Copy the outpost.conf
file provided in the top-level into /etc/init
.
$ cp /usr/local/factual/outpost/outpost.conf /etc/init/outpost.conf
Edit your config file
Edit the config.yml
file, and add in your Organization ID under the comment “the organization-id for the user”. Your account manager can provide this value upon request.
# the organization-id for the user
b3bda92e-b4b5-4012-bccc-48017cdefc81:
# the credentials for the organization, available via http://factual.com
api-key: ""
secret-key: ""
Next configure the values for source
that you will pass in to On-Prem. source
is the data source/exchange on which the impression originated. Your config.yml
file supports aliasing of data sources. This means that you can map the name of the exchange you pass to On-Prem, to the name that Foursquare uses to refer to that exchange.
The following table provides a mapping of these values:
Data Source/Exchange | Foursquare Source Value |
---|---|
MoPub | mopub |
AOL One | nexage |
Smaato | smaato |
Aerserv | aerserv |
OpenX | openx |
Factual | factual_dap |
Per the snippet below, setup your config.yml
with your desired data source mappings.
global:
directory: /tmp/on-prem/test-org
test:
audience:
sets:
max-memory: 1
fields: [designId, targetingCode, dataSource]
deployment-tags: []
data-source-aliases:
aerserv: internalAerservName
mopub: internalMoPubName
nexage: internalAOLName
smaato: internalSmaatoName
Using Upstart
We strongly recommend using Upstart. Under normal operation, On-Prem Server should never crash. However, if it should, you can take the simple precaution of using Upstart to help maintain availability in production.
You’ll need to confirm that your server can open a large number of file descriptors, which can be done with the following command:
$ limit nofile 500000 500000
Running the Server
Startup the On-Prem server by running the following command:
$ sudo start outpost
Once the server is running, the following message should appear in /var/log/factual/outpost/outpost.log
:
started HTTP server on 8989
If desired, the port can be changed in the outpost.conf
file. You can confirm that the server is running by hitting the /zz/health
endpoint.
$ curl http://localhost:8989/zz/health
ok
The server can be stopped and restarted with the following commands:
$ sudo stop outpost
$ sudo restart outpost
Testing Responses
Use curl for testing your installation. Remember to wrap the URL in quotes:
Audience test:
$ curl "http://[server]/geopulse/audience?user-id=<device_id>&source=<data_source>"
Proximity test:
$ curl "http://[server]/geopulse/proximity/indices?latitude=40.642875&longitude=-73.782991"
Match Response
Although Proximity and Audience are two different products, with their own specific use cases, On-Prem powers each, or both concurrently. To facilitate integration and onboarding, we’ve designed the response packets to be identical:
Attribute | JSON Field | Description |
---|---|---|
Design Name | designName | Name of the Proximity or Audience Design, assigned in the Designer UI. |
Design ID | designId | Foursquare-generated unique ID for the Design. |
Deployment ID | deploymentId | Foursquare-generated unique ID for the associated deploy of the Design. |
Deployment Tags | deploymentTags | String tags associated with the deploy of the Design as assigned. |
Targeting Code | targetingCode | A string assigned to the Set when designing the Design in the Designer UI. (In deprecated versions this was formerly called the “Group ID”.) |
Data Source | dataSource | Foursquare-provided ID for the associated audience data source. (Audience only) |
Metadata | metadata | A map of keys and values describing the matched set within the Design. |
Payload | payload | JSON payload describing the matched set, available as string or bytes (Proximity only) |
See the page on Terminology for detailed descriptions of each return value.
Proximity Response Example:
[
{
"designName": "BigBox",
"targetingCode": "Costco",
"deploymentId": "ABCEHGEEG3q45hasdShe",
"setId": "xyzHGEEG3q45hasdShe"
},
{
"designName": "Electronics",
"targetingCode": "Best Buy",
"deploymentId": "ABMnOwHGbdcG3q45haoazq",
"setId": "xyzHGEqa68Dxhasdhe"
}
]
Audience Response Example:
[
{
"designName": "affluent-business-travelers",
"designId": "17914b5c-899b-434a-b343-cbadec91e7ca",
"targetingCode": "affluentBiz",
"deploymentId": "AyoLwbEHGbdcG3q45haoae",
"setId": "xyzHlQbiqa353q45ha3h9e"
},
{
"designName": "san-diego-moms",
"designId": "9653e9dd-236d-4c53-8398-ca73a3876343",
"targetingCode":"SanDiegoMoms",
"deploymentId": "ABCEHGbdcGwZqb3q45hao9e8",
"setId": "xyzHqGExqsa353q45hys5he7"
}
]
Updated 8 months ago