hammoaj / node-red-contrib-cloudantplus Goto Github PK

/modules/shared/cloudantbase.js:292

Looks like the correct call is bulkDocuments().

The function sendDocumentOnPayload function has been copied verbatim within the following functions:

CloudantOutNode
https://github.com/hammoaj/node-red-contrib-cloudantplus/blob/master/cloudantplus.js#L223

and

CloudantInNode
https://github.com/hammoaj/node-red-contrib-cloudantplus/blob/master/cloudantplus.js#L467

I was going to raise a PR but wasn't sure if this was a stylistic choice or not.

Feel free to close if this is by design.

Passing an array in msg.payload is not resulting in a bulk document insert. Instead a single document is inserted with properties '0', '1', '2', etc. The node is configured for the insert operation, and to only take the msg.payload. I have also tried passing the array as the msg and unchecked the payload only option with no luck.

Currently an invalid database connection (either not found or invalid credentials) will crash Node-RED. This is due to the recursive way that connections are retried which sets a timer to re-call the connection request, and in doing so loses the Promise chain so the error is not trapped. If max retries is set to 1, then it does work as designed as the recursive call is never made and the Promise chain remains intact.

Investigating how to re-code to ensure that the retry attempts still return the correct Promise response back to the original call.

Problem code is in https://github.com/hammoaj/node-red-contrib-cloudantplus/blob/master/modules/shared/cloudantbase.js in the connecWithRetry function.

@Stwissel any ideas how this could be restructured to work?

I'd like to report an interesting issue while working in a Docker context. Although, it may be worth mentioning that I'm a Node-RED newbie so there may be an easy solution for my problem.

My Node-RED Dockerfile looks like this:

FROM nodered/node-red-docker
RUN npm install node-red-contrib-pdf-hummus
RUN npm install node-red-contrib-cloudantplus

My docker-compose.yml looks like this:

version: '3'

services:
  my-pipeline:
    build: .
    ports:
     - "1880:1880"
    volumes:
     - ./data/flows.json:/data/flows.json
    links:
     - cloudant-developer
  cloudant-developer:
    image: ibmcom/cloudant-developer:latest
    hostname: cloudant.dev
    ports:
      - "8080:80"
    volumes:
      - cloudant-developer:/srv

volumes:
  cloudant-developer:

When I run docker-compose up, I can see from the Node-RED logs the following issue:

my-pipeline_1 | 19 Nov 23:10:43 - [error] [cloudantplus out:insert] socket hang up

Now, there are few things I can't explain:

1. my flow contains an inject node but the option for firing at startup is not ticked in the node configuration. So, why is the cloudantplus out node triggered at startup?
2. after that, if I manually trigger the flow using the inject node I get the same issue reported during startup (i.e. "socket hang up") but if I make a non significant change (e.g. moving some node) in order to enable the Deploy button and then I actually re-deploy, everything works perfectly. Why a re-deploy of the exact same flow is required in order to fix the issue?

As a further note, I've tried to run the same images without docker-compose by starting first Cloudant Developer and then Node-RED, giving some time to Cloudant to startup and everything worked as expected.

Ultimately, it seems that the issue is caused by the cloudant out node trying to connect to Cloudant before the latter is actually up and running. But the worst thing is that, after the first attempt, the cloudant out node remains in a broken state until another deploy is triggered.

Is there something that can be "fixed" in the cloudant out node or the common approach is to ping some health endpoint of cloudant as suggested in the Docker documentation?

I installed Node-RED v2.2.2 on macOS and the cloudantplus nodes.
Observations:

• Node Cloudantplus does not appear in the palette

• When I configure the properties I'm prompted for user and password. On IBM Cloud I only have username, apikey etc.

I saw this tutorial, maybe it is not up-to-date (IAM?). Is there a solution?

Hi i like your project (https://flows.nodered.org/node/node-red-contrib-cloudantplus), but when i try to connect NR and Cloudant via this tutorial, i have this error (photo):
https://cloud.ibm.com/docs/Cloudant?topic=Cloudant-creating-an-ibm-cloudant-instance-on-ibm-cloud

Can you help me, please?