NCD local gateway simulator

Local gateway simulator to stress test NCD.

Available as Docker image instedd/lgwsim

Usage

Create executable by running:

shards build --release

Run:

bin/lgwsim

Usage with InSTEDD cloud

You need to create a QST server channel in order to use this application. The easiest way is to simulate a lgw registration.

1. Get a 4-digit registration code

$ curl -d "address=12341234" https://nuntium-stg.instedd.org/tickets.json
{"code":"5244","secret_key":"32a2a8ed-****-****-****-************","status":"pending","data":{"address":"12341234"}}

2. Create a QST server (local gateway) from an InSTEDD Application

3. Enter the 4-digit registration code: 5244

4. Get the generated

$ curl "https://nuntium-stg.instedd.org/tickets/5244.json?secret_key="32a2a8ed-****-****-****-************""
{"code":"5244","secret_key":"32a2a8ed-****-****-****-************","status":"complete","data":{"address":"12341234","channel":"my_qst_simulator","account":"manas","password":"Ji******","message":null}}

5. Lauch a QST Simulator with the above configuration

$ docker run --rm -it -e HOST=nuntium-stg.instedd.org -e ACCOUNT=manas -e CHANNEL_NAME=my_qst_simulator -e CHANNEL_PASSWORD=Ji****** instedd/lgwsim

6. Start sending messages from the InSTEDD Application

Example using variables: docker run --rm -it -e HOST=nuntium-stg.instedd.org -e ACCOUNT=email@email.com -e CHANNEL_NAME=my_channel -e CHANNEL_PASSWORD=password -e DELAY_REPLY_MIN_SECONDS=20 -e DELAY_REPLY_MAX_SECONDS=25 -e DELAY_REPLY_PERCENT=1 instedd/lgwsim

Usage in you local environment

If you are running Nuntium in your local environment, follow the steps above to create a QST server. Your config.yml should probably have these two parameters as follows:

host: web.nuntium.lvh.me port: 80

Behaviour

For each message received, if a reply is sent, it will be according to de following rules

Input	Reply
#oneof:A,B,C	A or B or C
#numeric:N-M	a number between N and M

For creating questionnaire steps these rules should be in the SMS message, make sure the responses are configured accordingly

The following environment variables controls other aspects of the behavior

Variable	Default	Description
SLEEP_SECONDS	10	How much time to wait between fetches from the QST server (Nuntium will mark the channel as down if it's not polled every 2 minutes)
NO_REPLY_PERCENT	0.2	Percent of respondents that never reply
DELAY_REPLY_PERCENT	0.2	Percent of respondents that have a delay in their reply
DELAY_REPLY_MIN_SECONDS	0	Of the above, minimum time in seconds of that delay (min..max)
DELAY_REPLY_MAX_SECONDS	60	Of the above, maximum time in seconds of that delay (min..max)
INCORRECT_REPLY_PERCENT	0.2	Percent of respondents that reply an incorrect answer
STICKY_RESPONDENTS	true	If true, once a respondent replies, it will always reply
INCOMING_BATCH_SIZE	100	How many incoming messages to fetch at once