Is your feature request related to a problem? Please describe.
Filling out the container name on the container node (and other nodes that use it) is a bit tedious. I have to constantly go check if the container name is correct.

Describe the solution you'd like
The container name field can be an autocomplete field instead that queries Node-RED for available docker container names. This would be a nice quality of life improvement for this Node-RED module.

Describe the bug
After upgrade to Node-RED 3.0 all Container Exec nodes fail with "System Error: [undefined] undefined"

To Reproduce
Steps to reproduce the behavior:

1. Upgrade Node-RED to 3.0
2. Create a new Container Exec flow
3. Inject the flow
4. Debug log will show System Error: [undefined] undefined

Expected behavior
Working fine before 3.0

Screenshots
If applicable, add screenshots to help explain your problem.

Additional context
Add any other context about the problem here.

Is your feature request related to a problem? Please describe.
I'm using dockerode to run a curl command and the output of the curl command is being streamed forcing me to use a join node to get the full response back. The join node works well but I have to use a timeout of 1 second to wait for the full response to come back. This just ends up adding extra latency to the whole automation.

Describe the solution you'd like
It would be nice to have an option for container run so that the output is returned as one whole string instead of streaming it.

Describe alternatives you've considered
Using the join node to get the full response with the caveat that it adds latency to wait for all messages

Hello,
I love your project, it works very well when I use it to get info from the Docker host or run simple commands (inspect, list, remove etc) but I have troubles with the update command.
I loaded the example flow and also in that when I run Node>Update it throws the error "Sytem Error: [400] undefined"
For example I need to set a Label for a node, so it should be docker node update "node" --label-add "label",
I have really no idea how to pass the missing parameters to the function.
Any help will be really appreciated.
Thank you.

Describe
Hello. I have a daily routine that runs a container, but sometimes there is a failure and the container is not executed, then shortly after Node-RED restarts. It's not something that happens all the time, just occasionally. If more information is needed, I can try to gather it.

Tag
node-red-contrib-dockerode 0.12.1

The 'Docker Run' node is configured as shown in the screenshot.

Below are the logs from when the error occurred:

2024-04-10T08:31:15.864932772Z /opt/service/build/nodeRedUserDir/node_modules/docker-modem/lib/modem.js:405
2024-04-10T08:31:15.864996077Z   stream.on('data', onStreamEvent);
2024-04-10T08:31:15.865014605Z          ^
2024-04-10T08:31:15.865031942Z
2024-04-10T08:31:15.865048218Z TypeError: Cannot read property 'on' of null
2024-04-10T08:31:15.865064661Z     at Modem.followProgress (/opt/service/build/nodeRedUserDir/node_modules/docker-modem/lib/modem.js:405:10)
2024-04-10T08:31:15.865081104Z     at /opt/service/build/nodeRedUserDir/node_modules/node-red-contrib-dockerode/dist/docker-container-actions.js:116:50
2024-04-10T08:31:15.865097630Z     at /opt/service/build/nodeRedUserDir/node_modules/dockerode/lib/docker.js:118:7
2024-04-10T08:31:15.865114230Z     at /opt/service/build/nodeRedUserDir/node_modules/docker-modem/lib/modem.js:331:7
2024-04-10T08:31:15.865130602Z     at IncomingMessage.<anonymous> (/opt/service/build/nodeRedUserDir/node_modules/docker-modem/lib/modem.js:350:9)
2024-04-10T08:31:15.865147919Z     at IncomingMessage.emit (events.js:326:22)
2024-04-10T08:31:15.865164416Z     at IncomingMessage.EventEmitter.emit (domain.js:483:12)
2024-04-10T08:31:15.865199627Z     at endReadableNT (_stream_readable.js:1241:12)
2024-04-10T08:31:15.865213307Z     at processTicksAndRejections (internal/process/task_queues.js:84:21)
2024-04-10T08:31:16.542568296Z npm ERR! code ELIFECYCLE
2024-04-10T08:31:16.544732816Z npm ERR! errno 1
2024-04-10T08:31:16.628621246Z npm ERR! [email protected] start: `node index.js`
2024-04-10T08:31:16.628659608Z npm ERR! Exit status 1
2024-04-10T08:31:16.628674622Z npm ERR!
2024-04-10T08:31:16.628688115Z npm ERR! Failed at the [email protected] start script.
2024-04-10T08:31:16.628701601Z npm ERR! This is probably not a problem with npm. There is likely additional logging output above.
2024-04-10T08:31:16.718690847Z
2024-04-10T08:31:16.718804111Z npm ERR! A complete log of this run can be found in:
2024-04-10T08:31:16.718861075Z npm ERR!     /root/.npm/_logs/2024-04-10T08_31_16_605Z-debug.log
2024-04-10T08:31:19.836267028Z

Can we get support for passing the docker.sock file in? I would prefer this over exposing my docker on the network.

I've setup docker to run on TCP with TLS, because doing without wouldn't let me start docker.
Now, neither in Node Red or in the Docs I see an option to configure dockerode with TLS.

Am I overseeing something or is it not possible doing this? Or is there another way I have to configure docker to be useable without TLS?

Is your feature request related to a problem? Please describe.
I would love to only pull the first stats of a container.

Describe the solution you'd like
Add a "no stream" option to the container stats node

Additional context
Docker supports a --no-stream option for container stats. Documentation

The implemantation would look something like this:

  container.stats({ stream: false }, function (err, data) {
      if (err) {
          node.error('Error:', err)
      }

      node.send(data)
  });

Hello, was trying to get the logs of a running container, I connected using the socket method and could successfully get the docker events and see things like container inspect and container top. But when I try to do a container logs I get:

System Error:  [400] undefined

in the debug log, which seems to come from line 278 of docker-container-actions.js which came from:

                            // https://docs.docker.com/engine/api/v1.40/#operation/ContainerLogs
                            container.logs()

not sure how to fix this... changing it to:

                            container.logs({
		                          stdout: true,
		                          stderr: true
	                          })

now returns a buffer. Every invocation returns the same buffer though!
Further changing to:

                            let since = node.context().get("log_since") || 0;
                            node.warn("Getting logs since "+since);
                            container.logs({
		                          stdout: true,
		                          stderr: true,
                              since: since
	                          })
                                .then(function (res) {
                                  node.status({ fill: 'green', shape: 'dot', text: containerId + ' restarted' });
                                  node.context().set("log_since",Math.floor(+new Date() / 1000));
                                node.send(Object.assign(msg, { payload: res }));
                            }).catch(function (err) {

stores the last time the logs were grabbed, and upon calling again just gives the new log entries... Could I ask for this please?

Hello,
I want to design an application to create and manage containers. This should be done in Node-RED, because the Conatiner, which should be managed, are itself based on Node-RED.
My question now is how to create a container with node-red-contrib-dockerode. Do the options to create have to be packed into the incoming payload of the nodes?

Up to now I have stored the function in a shell script. In principle I want to realize everything from the shell script in node-RED. The script I state with a number to enumerate the adapters.

#!./bin/bash
$1
number=$1
containername=adapter_modbus$number
volumename=vol_adapter_modbus$number
porthost=$((55500 + $number))
adapterip=$((100 + $number))
docker volume create $volumename
docker create --net adapter_net --ip 172.20.0.$adapterip -p $porthost:1880 -v $volumename:/data --name $containername node-red_modus_image
docker start $containername

I would have thought this would be an available command in the Image node, but cannot find it. Nor in the Container node. I've looked through all the others and can't seem to see anything obvious, but may have missed something.

I'm looking to run the following in node-red:

docker run -it --rm --name certbot certbot/dns-linode renew -q

This isn't achievable with an exec node without configuring access beyond the container. I've a certbot docker image and have docker.sock configured and accessible in node-red. I was hoping node-red-contrib-dockerode might have a means to achieve this.

Hi,

Thanks for your work.. It looks really nice.

I wanted to play around with it to see what are the possibilities, so the logical thing would be to check the examples, however the types are unknown although I have installed it via the "manage pallette" in node-red. THe message is "unknown-type" for all items such as docker-swarm or whatever type.

Thank you for your help!!

Describe the bug
I am just using an Service List node with no service mentioned, as it should be the case, since I want to list the available services:

I did the exact same for the Container List node, which works fine:

Even when entering the name or ID of a service just to test the result, I get the same error.

Here is the full traceback:

TypeError: Cannot read property 'options' of undefined
    at DockerServiceAction._inputCallback (/data/node_modules/node-red-contrib-dockerode/dist/docker-service-actions.js:19:67)
    at /usr/src/node-red/node_modules/@node-red/runtime/lib/nodes/Node.js:210:26
    at Object.trigger (/usr/src/node-red/node_modules/@node-red/util/lib/hooks.js:166:13)
    at DockerServiceAction.Node._emitInput (/usr/src/node-red/node_modules/@node-red/runtime/lib/nodes/Node.js:202:11)
    at DockerServiceAction.Node.emit (/usr/src/node-red/node_modules/@node-red/runtime/lib/nodes/Node.js:186:25)
    at DockerServiceAction.Node.receive (/usr/src/node-red/node_modules/@node-red/runtime/lib/nodes/Node.js:485:10)
    at Immediate._onImmediate (/usr/src/node-red/node_modules/@node-red/runtime/lib/flows/Flow.js:831:52)
    at processImmediate (internal/timers.js:464:21)

To Reproduce
Steps to reproduce the behavior:

1. Configure the module
2. Add a Service List node
3. Start the node with an Input node
4. See error

Expected behavior
This should list the services.

Additional context
I just want to list all containers, check if one is unhealthy, and restart its service if so.

https://flows.nodered.org/node/node-red-contrib-dockerode

I used your github.com npm publish tool chain....

The new version shows up here:
https://github.com/naimo84/node-red-contrib-dockerode
https://www.npmjs.com/package/node-red-contrib-dockerode

Every time on Deploy I get "Not properly configured" message for the node which is working just fine.

I'm using this node to restart my mosquitto container:
[{"id":"e5c962cd3ec8f581","type":"docker-container-actions","z":"ea6ed994.f39728","name":"MQTT","config":"c0b27d850da8e8ba","container":"mosquitto","containertype":"str","action":"restart","options":"","optionstype":"str","image":"","imagetype":"str","pullimage":false,"deletecontainer":false,"stream":false,"createOptions":"","startOptions":"","createOptionsType":"json","startOptionsType":"json","x":1190,"y":80,"wires":[[]]},{"id":"c0b27d850da8e8ba","type":"docker-configuration","host":"/var/run/docker.sock","port":"","ca":"","cert":"","key":""}]

It shows problems with createOptions and startOptions

which I don't need, I'm just restarting it (and it indeed restarts)

Really annoying every time I make a deploy :)

Thanks

Hello, I using 2 host, one for docker and one for node-red. When the docker host restart the event node say that he is disconnected but do not reconnect after Docker is back up. If I deploy again the event node connect like expected.
Can I force a reconnexion or do something for this ?
Thanks

Is your feature request related to a problem? Please describe.
When pulling logs from a container if you do a re-deploy it will pull down the full log. This is a problem for me.

The issue is that I needed the logs for a container saved to a file locally so fail2ban could parse it and do it's thing. I noticed whenever I re-deploy in Node-RED it would send the full log contents again, writing it to my file, and making fail2ban think there was a bunch of new login attempts because they are new rows.

Describe the solution you'd like
I would like to be able to pass in msg.since and if this is present on the message the node should use this instead of it's own internal since property.

Honestly though I would like to see all of the options exposed via message variables. This way I can grab all log rows between two dates for example.

Also the node outputs a debug message with how many messages it is grabbing from since which is a bit annoying. I would like a checkbox on the node called output debug info that if unchecked will not do any sort of debugging. I'm pulling logs every 2 seconds so this message spams my node-red GUI (and I know I can silence it, but the messages honestly only need to be enabled for debugging anyways).

Describe alternatives you've considered
What I did to fix this for now is just ignore the first response back from the container logs on deploy.

Can we get descriptions for these nodes similar to what other people do with their modules? It would be really nice to know all the fields I can set and that get returned without having to dig into the code.

Currently the description section is just blank.

Thank you for this node-red contrib.

Is your feature request related to a problem? Please describe.
Delete a container after run it

Describe the solution you'd like
When action = run, add a checkbox "delete container after run"

Describe alternatives you've considered
docker events (where is the container id ? Actor.ID or Actor.Attributes.container ? + disconnect + container delete => no error but container not deleted

Additional context
Add any other context or screenshots about the feature request here.

Hi @ethanbrooks,

Many many thanks for your contribution on my nodes. I was so cheeky and merged your fork back. The name node-red-contrib-dockerode was on purpose, because of the main library dockerode.

If you have further changes, feel free to create a new pull request. I' merge it as soon as possible.

Greets from Germany,
Benjamin

Currently these nodes override whatever the message had when it came into the node. This should actually keep the original message and just set the keys/values on that message object before returning it.

A good example of why this is not the correct way is if you wire up HTTP request node before and a HTTP response node after one of the dockerode node's it wont be able to send the response because the msg.response is not defined (it was removed by the dockerode node).

Giving direct access to the docker.sock seems to be frowned upon from security perspective.

Does this node support using a docker proxy instead of accessing docker.sock directly?

Describe the bug
This package should be validating that the user accessing any http routes has flows.write permission but it currently does not. This means anyone that knows the Node-RED instance URL can hit these endpoints.

This is mitigated a bit by the fact the attacker would need the config node's id in order to make a successful request. People do share config flows online though and often times the id field still references the original id of the config node.

To Reproduce
Steps to reproduce the behavior:

1. Make a POST request to <instance>/containerSearch with payload {"id": ""}
2. You get an error TypeError: Cannot read properties of null (reading 'getClient') (unless you passed a valid ID then you get the full payload back)

Expected behavior
Should deny permission.

Screenshots
N/A

Additional context
Only reason I found this was because I was digging into the source to fix other issues. This definitely needs to be resolved.

It would also be good to move all routes under a sub-directory such as /docker to avoid collisions with other packages or user defined routes.

Is your feature request related to a problem? Please describe.
I want to handle errors that occur on the dockerode nodes.

Describe the solution you'd like
Any error should trigger the Catch node of nodered.

Additional context
Calling node.error() with one argument will not trigger a Catch node.
So instead of calling node.error(err) the function should be called with the original message as the second argument.
E.g.: node.error(err, msg)
Documentation

This had been working for awhile, but now I have this error when I try to inspect a container.

also: [error] [docker-container-actions:7d64bf1b.df5b] Sytem Error: [undefined] undefined when i try to restart a container

node-red-contrib-dockerode
0.4.21

inside docker: Docker version 20.10.8, build 3967b7d28e

docker-compose:

version: "2.2"
services:
  node-red:
    container_name: node-red
    restart: unless-stopped
    network_mode: host
    image: nodered/node-red:latest
    volumes:
      - /var/run/docker.sock:/var/run/docker.sock
      - /home/remote_fs_user/conf/node_red/data:/data
      - /home/remote_fs_user/conf/home-assistant/tensorflow:/tensorflow
    environment:
      - TZ="Europe/Zurich"
      - PROMETHEUS_COLLECT_DEFAULT_METRICS="true"
    user: 1000:1000
    group_add: [977]
    depends_on:
      - mosquitto
      - home-assistant

logs:

Welcome to Node-RED
===================

7 Oct 08:28:33 - [info] Node-RED version: v2.0.6
7 Oct 08:28:33 - [info] Node.js  version: v14.17.6
7 Oct 08:28:33 - [info] Linux 5.10.69-1-lts x64 LE
7 Oct 08:28:33 - [info] Loading palette nodes
7 Oct 08:28:35 - [info] Settings file  : /data/settings.js
7 Oct 08:28:35 - [info] Context store  : 'default' [module=memory]
7 Oct 08:28:35 - [info] User directory : /data
7 Oct 08:28:35 - [warn] Projects disabled : editorTheme.projects.enabled=false
7 Oct 08:28:35 - [info] Flows file     : /data/flows.json
7 Oct 08:28:35 - [info] Server now running at http://127.0.0.1:1880/
7 Oct 08:28:35 - [info] Starting flows
7 Oct 08:28:35 - [info] Instanciating PrometheusMetricConfigNode shield_source
7 Oct 08:28:35 - [info] Initializing Prometheus exporter
7 Oct 08:28:35 - [info] Prometheus metrics are available at path /metrics
7 Oct 08:28:35 - [info] PrometheusMetricConfigNode.registerMetric shield_source
7 Oct 08:28:35 - [info] Added Prometheus Counter shield_source
7 Oct 08:28:35 - [info] Instanciating PrometheusExporterNode 8fd3c501.be3b28
7 Oct 08:28:35 - [info] Started flows
7 Oct 08:28:35 - [info] [tcp in:cf78a798ea8a2924] listening on port 1113
7 Oct 08:28:35 - [info] [server:Home Assistant] Connecting to http://192.168.2.24:8123
7 Oct 08:28:35 - [info] [mqtt-broker:Fractal] Connected to broker: Node-red@mqtt://192.168.2.24:1883
7 Oct 08:28:35 - [info] [server:Home Assistant] Connected to http://192.168.2.24:8123
7 Oct 08:28:55 - [error] [docker-container-actions:Inspect container] TypeError: Cannot read property 'comand' of undefined
(node:17) UnhandledPromiseRejectionWarning: TypeError: Cannot read property 'error' of undefined
    at /data/node_modules/node-red-contrib-dockerode/dist/docker-container-actions.js:701:50
    at runMicrotasks (<anonymous>)
    at processTicksAndRejections (internal/process/task_queues.js:95:5)
(Use `node --trace-warnings ...` to show where the warning was created)
(node:17) UnhandledPromiseRejectionWarning: Unhandled promise rejection. This error originated either by throwing inside of an async function without a catch block, or by rejecting a promise which was not handled with .catch(). To terminate the node process on unhandled promise rejection, use the CLI flag `--unhandled-rejections=strict` (see https://nodejs.org/api/cli.html#cli_unhandled_rejections_mode). (rejection id: 1)
(node:17) [DEP0018] DeprecationWarning: Unhandled promise rejections are deprecated. In the future, promise rejections that are not handled will terminate the Node.js process with a non-zero exit code.
7 Oct 08:35:22 - [info] Stopping flows
7 Oct 08:35:22 - [info] [tcp in:cf78a798ea8a2924] stopped listening on port
7 Oct 08:35:22 - [info] [server:Home Assistant] Closing connection to http://192.168.2.24:8123
7 Oct 08:35:22 - [info] Closing PrometheusMetricConfigNode shield_source
7 Oct 08:35:22 - [info] Removed Prometheus metric shield_source
7 Oct 08:35:22 - [info] AlexaLocalNode closing done...
7 Oct 08:35:22 - [info] [mqtt-broker:Fractal] Disconnected from broker: Node-red@mqtt://192.168.2.24:1883
7 Oct 08:35:22 - [info] Stopped flows
7 Oct 08:35:22 - [info] Starting flows
7 Oct 08:35:22 - [info] Instanciating PrometheusMetricConfigNode shield_source
7 Oct 08:35:22 - [info] PrometheusMetricConfigNode.registerMetric shield_source
7 Oct 08:35:22 - [info] Added Prometheus Counter shield_source
7 Oct 08:35:22 - [info] Instanciating PrometheusExporterNode 8fd3c501.be3b28
7 Oct 08:35:22 - [info] Started flows
7 Oct 08:35:22 - [info] [tcp in:cf78a798ea8a2924] listening on port 1113
7 Oct 08:35:22 - [info] [server:Home Assistant] Connecting to http://192.168.2.24:8123
7 Oct 08:35:22 - [info] [mqtt-broker:Fractal] Connected to broker: Node-red@mqtt://192.168.2.24:1883
7 Oct 08:35:22 - [info] [server:Home Assistant] Connected to http://192.168.2.24:8123
7 Oct 08:35:24 - [error] [docker-container-actions:Inspect container] TypeError: Cannot read property 'comand' of undefined
7 Oct 08:35:27 - [error] [docker-container-actions:Inspect container] TypeError: Cannot read property 'comand' of undefined
7 Oct 08:36:07 - [error] [docker-container-actions:Inspect container] TypeError: Cannot read property 'comand' of undefined

Describe the bug
When you create a new container node and configure it, it shows that the config is invalid (see screenshot below)

To Reproduce
Drag over a new node, configure it correctly and the error will show when you try to deploy.

Expected behavior
The node is configured correctly so it should work

Screenshots

Additional context
I can get around this by exporting the node and setting both of these options to "{}" which is a bit of an annoying workaround.