I have an endpoint inside a docker container. When I call the endpoint from postman it works as intended. When I call it from node.fetch from another container it also works but when I call it from another docker container using chai, I get the following error:
curl: (7) Couldn't connect to server
The chai call works well, since I can see the correct logs. How ever when I get to the following line of code it doesn't behave as expected.
This is the chai code:
res = await chai
.request(server)
.post('/bot_manager/management/initiliaze')
.set('content-type', 'application/json')
.send(body)
This is the line where it fails:
exec(`curl --unix-socket /var/run/docker.sock -H "Content-Type: application/json" -d '{ "Image": "strategy_baseline", "ExposedPorts": { "${PORT}/tcp": {} }, "HostConfig": { "Binds": ["utils:/usr/src/app/utils:delegated","${process.env.CURRENT_PATH}/src/bot_manager/api/strategies/${botId}.js:/usr/src/app/strategies/${botId}.js"], "NetworkMode": "goatFish_backend", "PortBindings": { "${PORT}/tcp": [{ "HostPort": "${PORT}" }]}}, "Env": ["BOTNAME=${botId}", "PORT=${PORT}", "PAIR=${PAIR}"]}' -X POST http:/v1.4/containers/create?name=${botId}`, (err, stdout, stderr) => {// empty}
This works well in every method except chai. I also use the exact sam input through out the different methods.
Any help or indications are much appreciated. TIA!!
Related
I've googled the f out of this.
here is the log driver config for my docker container in the compose file
driver: gelf
options:
gelf-address: "http://graylog:12201"
I created a GELF HTTP input in the admin console.
I know that graylog is accessible at 12201, because if I ssh into a container and run
curl -X POST -H 'Content-Type: application/json' -d '{ "version": "1.1", "host": "example.org", "short_message": "A short message", "level": 5, "_some_info": "test" }' 'http://graylog:12201/gelf'
Then I can see the log message.
The problem is, it seems like I have to add the /gelf to the address, but docker complains if I try to do that. But no other curl commands work without it, and I can't seem to get it to work with TCP or UDP at all. So.... what am I doing wrong?
I started my node container with this flags:
daemon=1
printtoconsole=1
testnet=1
rpcport=9332
rpcallowip=0.0.0.0/0
rpcuser=user
rpcpassword=password
rpcbind=0.0.0.0
server=1
I opened port in my docker-compose :
node:
image: bitcoin-sv
container_name: 'node'
restart: always
ports:
- '9332:9332'
I can call methods from bitcoin-cli in my container
docker exec -it node bash
root#9196d074e4d8:/opt/bitcoin-sv# ./bitcoin-cli getinfo
But I cannot call it from curl
curl --user user --data-binary '{"jsonrpc": "1.0", "id":"curltest", "method": "getinfo, "params": ["", 0.1, "donation", "seans outpost"] }' -H 'content-type: text/plain;' http://127.0.0.1:9332
Enter host password for user 'user':
curl: (52) Empty reply from server
How can i call it from curl? Maybe i have to call to cli?
not sure what can be your issue, but the first approach would be do the curl inside the container in order to verify that the HTTP interface is working properly. So you should try this:
docker exec -it node bash
root#9196d074e4d8:/opt/bitcoin-sv# curl --user user --data-binary '{"jsonrpc": "1.0", "id":"curltest", "method": "getinfo, "params": ["", 0.1, "donation", "seans outpost"] }' -H 'content-type: text/plain;' localhost:9332
Once you are sure the interface is working inside the container, you can move forwred and try it from the host.
I'm using IoTAgent over MQTT configuration. After going step by step with the tutorial presented on FIWARE's website is there a way to invoke command using measurment input of the IoT Agent?
Let's say i have 2 arduinos: one is an actuator and the other one is a sensor. The actuator has LED attached, sensor has a button. I want to send a message from sensor-arduino with command ON (to MQTT Broker or directly as an Ultralight message via HTTP - as far as I tested IoTA for Ultralight can run simultaneously both modes which is great) what will invoke sending the command defined for given device.
Let's say im using this config:
curl -iX POST \
'http://localhost:4041/iot/devices' \
-H 'Content-Type: application/json' \
-H 'fiware-service: openiot' \
-H 'fiware-servicepath: /' \
-d '{
"devices": [
{
"device_id": "bell001",
"entity_name": "urn:ngsi-ld:Bell:001",
"entity_type": "Bell",
"protocol": "PDI-IoTA-UltraLight",
"transport": "MQTT",
"commands": [
{ "name": "ring", "type": "command" }
],
"static_attributes": [
{"name":"refStore", "type": "Relationship","value": "urn:ngsi-ld:Store:001"}
]
}
]
}
'
I can invoke command like this (which is very inconvenient):
curl -iX POST \
'http://localhost:4041/v1/updateContext' \
-H 'Content-Type: application/json' \
-H 'fiware-service: openiot' \
-H 'fiware-servicepath: /' \
-d '{
"contextElements": [
{
"type": "Bell",
"isPattern": "false",
"id": "urn:ngsi-ld:Bell:001",
"attributes": [
{ "name": "ring", "type": "command", "value": "" }
],
"static_attributes": [
{"name":"refStore", "type": "Relationship","value": "urn:ngsi-ld:Store:001"}
]
}
],
"updateAction": "UPDATE"
}'
Or after registering the command, I can use Orion Context Broker:
curl -iX PATCH \
'http://localhost:1026/v2/entities/urn:ngsi-ld:Lamp:001/attrs' \
-H 'Content-Type: application/json' \
-H 'fiware-service: openiot' \
-H 'fiware-servicepath: /' \
-d '{
"on": {
"type" : "command",
"value" : ""
}
}'
Those approaches give me a response in Mosquitto subscriber.
How can I create a message sent to IoTAgent (via MQTT or HTTP) that will invoke command sent to MQTT Broker? The command is further managed in actuator-arduino as long as is received in the MQTT Broker.
Using the IoT Agent North Port
I can invoke command like this (which is very inconvenient):
This is a command direct to the North Port of the IoT Agent itself - as the tutorial states it should only be used for testing the connectivity. You should never need to do this yourself - this is the command the Orion Context Broker sends to the IoT Agent
Using NSGI v2
Or after registering the command, I can use Orion Context Broker:
Using the PATCH command is the way to go - pre-registration of commands is not necessary in the more recent releases of the IoT Agent Library.
How to use FIWARE to send Commands to Arduino
How can I create a message sent to IoTAgent (via MQTT or HTTP) that will invoke the command sent to MQTT Broker? The command is further managed in actuator-arduino as long as is received in the MQTT Broker.
The context broker is merely receiving changes of context and informing subscribed services. Within the tutorial an NSGI v2 PATCH request is programmatically sent to the Context Broker using an HTTP request :
https://github.com/FIWARE/tutorials.Step-by-Step/blob/master/context-provider/controllers/ultraLight.js
const options = {
method: 'PATCH',
url: UL_CONTEXT_BROKER + '/entities/' + UL_NGSI_PREFIX + id + '/attrs',
headers: {
'Content-Type': 'application/json',
'fiware-servicepath': '/',
'fiware-service': 'openiot'
},
body: payload,
json: true
};
request(options, error => {
if (error) {
debug(error);
}
});
With a configured IoT Agent the result will be a topic posted to the MQTT broker.
There needs to be some code in the device to ensure it is subscribed to the right topic and then receive the payload.
https://github.com/FIWARE/tutorials.Step-by-Step/blob/master/context-provider/iot.js
const mqtt = require('mqtt');
const apiKey = process.env.DUMMY_DEVICES_API_KEY || '1234';
const topics = '/' + apiKey + '/#';
const mqttBrokerUrl = process.env.MQTT_BROKER_URL || 'mqtt://mosquitto';
global.MQTT_CLIENT = mqtt.connect(mqttBrokerUrl);
MQTT_CLIENT.on('connect', () => {
debug('Subscribing to MQTT Broker: ' + mqttBrokerUrl + ' ' + topics);
MQTT_CLIENT.subscribe(topics);
MQTT_CLIENT.subscribe(topics + '/#');
});
mqtt.connect(mqttBrokerUrl);
MQTT_CLIENT.on('message', function(topic, message) {
// message is Buffer
Ultralight.processMqttMessage(topic.toString(), message.toString());
});
I am using Icinga Version 2.4.2 to monitor services on several hosts. I would like to be able to place certain hosts in maintenance mode for a set amount of time using a cli tool or rest API instead of the Web UI.
Is this possible and if so what tool/api should I use?
If I cannot do this through a remote tool/api what command should I use on the server or client to place clients in maintenance mode?
Update: It seems like the rest api has a solution. This set of permissions works:
object ApiUser "root" {
password = "foobar"
permissions = [ "console", "objects/query/Host", "objects/query/Service", "actions/schedule-downtime", "actions/remove-downtime"]
}
Then the following allows me to make and remove downtimes:
curl -k -s -u root:foobar -H 'Accept: application/json' -X POST "https://localhost:5665/v1/actions/schedule-downtime?filter=host.name==%22${TARGET}%22&type=Host" -d '{ "start_time": "1528239116", "end_time": "1528325561", "duration": 1000, "author": "root", "comment": "downtime on $TARGET" }' | jq .
curl -k -s -u root:foobar -H 'Accept: application/json' -X POST "https://localhost:5665/v1/actions/remove-downtime?filter=host.name==%22${TARGET}%22&type=Host" | jq .
Right now the only issue with this I am having is how to pass in variables for the start and stop dates. Attempting this keeps resulting in the following error:
{
"status": "Invalid request body: Error: lexical error: invalid char in json text.\n { \"start_time\": $current_time,\n (right here) ------^\n\n",
"error": 400
}
I'm new to docker. I have read the tutorial in docker remote API . In aspect of creating container. It show me too many param to fill. I want to know what is equivalent to this command :
docker run -d -p 5000:5000 --restart=always --name registry
registry:2.
I have no idea about it. Can anyone tell me? Thanks!
Original answer (July 2015):
That would be (not tested directly), as in this tutorial (provided the remote API is enabled):
First create the container:
curl -v -X POST -H "Content-Type: application/json" -d '{"Image": " registry:2.",}' http://localhost:2376/containers/create?name=registry
Then start it:
curl -v -X POST -H "Content-Type: application/json" -d '{"PortBindings": { "5000/tcp": [{ "HostPort": "5000" }] },"RestartPolicy": { "Name": "always",},}' http://localhost:2376/containers/registry/start?name=registry
Update February 2017, for docker 1.13+ see rocksteady's answer, using a similar idea but with the current engine/api/v1.26.
More or less just copying VonCs answer in order to update to todays version of docker (1.13) and docker remote api version (v1.26).
What is different:
All the configuration needs to be done when the container is created, otherwise the following error message is returned when starting the container the way VonC did.
{"message":"starting container with non-empty request body was deprecated since v1.10 and removed in v1.12"}
First create the container: (including all the configuration)
curl -v -X POST -H "Content-Type: application/json" -d #docker.conf http://localhost:2376/containers/create?name=registry
The file docker.conf looks like this:
{
"Image": registry:2.",
"ExposedPorts": {
"5000/tcp": {}
},
"HostConfig": {
"PortBindings": {
"5000/tcp": [
{
"HostPort": "5000"
}
]
},
"RestartPolicy": {
"Name": "always"
}
"AutoRemove": true
}
}
Then start it: (the parameter name is not necessary, the container is just named registry)
curl -v -X POST -H "Content-Type: application/json" http://localhost:2376/containers/registry/start
Create docker container in Docker Engine v1.24
Execute the post request -
curl -X POST -H "Content-Type: application/json" http://DOCKER_SERVER_HOST:DOCKER_PORT/v1.24/containers/create?name=containername
In the request body, you can specify the JSON parameters like
{
"Hostname": "172.x.x.x",
"Image": "docker-image-name",
"Volumes": "",
"Entrypoint": "",
"Tty": true
}
It creates your docker container
Start the container
Execute the POST request
curl -X POST http://DOCKER_SERVER_HOST:DOCKER_PORT/v1.24/containers/containername/start
Reference link - https://docs.docker.com/engine/api/v1.24/