CouchDB installation issue on Apple Silicon chip

Technical Support
42

@jkuester below is a screenshot of the terminal after running the command.

image
image1920×1011 158 KB

43

Failed to compute cache key: “/set-up-cluster.sh”: not found

Okay, so it looks like this line fails in the Dockerfile:

COPY --chown=couchdb:couchdb set-up-cluster.sh /opt/couchdb/etc/set-up-cluster.sh

Can you double-check to confirm that you do have a file at /Users/esthermoturi/cht-core/couchdb/set-up-cluster.sh?

Also can you let us know the result of docker -v? I have seen docker behave weirdly before when it comes to relative file paths. In this case, the Dockerfile has a relative reference to set-up-cluster.sh since it is just in the same directory as the Dockerfile.

44

This is the result , “Docker version 20.10.22, build 3a2c30b”

Can you double-check to confirm that you do have a file at /Users/esthermoturi/cht-core/couchdb/set-up-cluster.sh ?

The only file under couchDb is a Dockerfile

1 Like
45

@Esther_Moturi & @Kymoraa, @mrjones has just documented a simpler way for running the CHT in dev mode on Apple Silicon machines. I tested it this morning and it worked nicely:

Here are the steps:

node -v && npm -v
v20.11.0
10.2.4

In the directory of the cloned cht-core (latest master branch), I ran:

npm ci
npm run build-dev

docker build -t couchdb-apple-silicon couchdb/.
mkdir -p ~/cht-docker
curl -s -o ~/cht-docker/docker-compose.yml https://staging.dev.medicmobile.org/_couch/builds_4/medic:medic:master/docker-compose/cht-couchdb.yml
cat > ~/cht-docker/couchdb-override.yml << EOF
version: '3.9'
services:
    couchdb:
        image: couchdb-apple-silicon
        ports:
            - "5984:5984"
            - "5986:5986"
EOF

I then run the three npm run commands in the Developing section, and managed to have the CHT open at http://localhost:5988/ correctly.

@Esther_Moturi, when you have some time, can you please try this simplified approach?

1 Like
46

This is a problem! (And probably will cause issues with whatever approach you try…) Assuming your code is on a recent version, you should have quite a few files in that directory besides just the Dockerfile. Seems like something funky happened to your local cht-core. I would recommend deleting everything and re-cloning cht-core before trying the newly documented approach!

1 Like
47

You were right @jkuester , I got the files when I deleted and reinstalled cht-core

@andra , I am getting the error below when I run npm run dev-api and npm run dev-sentinel

[nodemon] app crashed - waiting for file changes before starting…

48

You were right Josh , I got the files when I deleted and reinstalled cht-core

@Esther_Moturi - did you successfully build the couchDB image per Andra’s comment above now that you have a complete CHT Core repo checked out?

I am getting the error below when I run npm run dev-api and npm run dev-sentinel

can you be specific which of the two command outputs those errors? Can you copy and paste in a more of the lines leading up to the error?

49

Yes, I did.

Can you be specific which of the two command outputs those errors?

Both commands. I ran them concurrently.

image
image1920×407 53 KB

image
image1636×532 158 KB

50

Hi @Esther_Moturi

The error is kinda buried in all that text, but it’s the same in both cases. It seems you didn’t define the COUCH_URL environment variable, which is required for both api and sentinel.

51

I do notice that our current docs only have the two calls to export for COUCH_NODE_NAME and COUCH_URL only for the CHT 3.x instructions. These are shared for both 3.x and 4.x should be moved up into the " CHT Core Cloning and Setup" above.

@Esther_Moturi try running this (below) and I’ll submit a PR to move those two calls per above:

echo "export COUCH_NODE_NAME=nonode@nohost">> ~/.$(basename $SHELL)rc
echo "export COUCH_URL=http://medic:password@localhost:5984/medic">> ~/.$(basename $SHELL)rc
. ~/.$(basename $SHELL)rc
52

Okay, now I am getting an error. So I ran the commands provided by @andra , then defined my couch_url and couch_node_name which once I echo I get the correct values as below

esthermoturi@Esthers-MacBook-Pro cht-core % echo $COUCH_NODE_NAME && echo $COUCH_URL
nonode@nohost
http://medic:password@localhost:5984/medic

I then run npm run dev-api and npm run dev-sentinel and get the error below continuously

RequestError: Error: connect ECONNREFUSED 127.0.0.1:5984
    at new RequestError (/Users/esthermoturi/cht-core/node_modules/request-promise-core/lib/errors.js:14:15)
    at plumbing.callback (/Users/esthermoturi/cht-core/node_modules/request-promise-core/lib/plumbing.js:87:29)
    at Request.RP$callback [as _callback] (/Users/esthermoturi/cht-core/node_modules/request-promise-core/lib/plumbing.js:46:31)
    at self.callback (/Users/esthermoturi/cht-core/node_modules/request/request.js:185:22)
    at Request.emit (node:events:527:28)
    at Request.onRequestError (/Users/esthermoturi/cht-core/node_modules/request/request.js:877:8)
    at ClientRequest.emit (node:events:527:28)
    at Socket.socketErrorListener (node:_http_client:462:9)
    at Socket.emit (node:events:527:28)
    at emitErrorNT (node:internal/streams/destroy:151:8) {
  cause: Error: connect ECONNREFUSED 127.0.0.1:5984
      at TCPConnectWrap.afterConnect [as oncomplete] (node:net:1229:16) {
    errno: -61,
    code: 'ECONNREFUSED',
    syscall: 'connect',
    address: '127.0.0.1',
    port: 5984
  },
  error: Error: connect ECONNREFUSED 127.0.0.1:5984
      at TCPConnectWrap.afterConnect [as oncomplete] (node:net:1229:16) {
    errno: -61,
    code: 'ECONNREFUSED',
    syscall: 'connect',
    address: '127.0.0.1',
    port: 5984
  }
}

Could this be happening because I have CouchDB2 installed?