DEV Community: cambiph

NPM run scripts exit status in Docker

cambiph — Mon, 07 Oct 2019 08:35:21 +0000

Hello all,

I am confused to how the exit code from a process inside a container gets passed to docker.
I have a container which runs an npm run script.
The npm run script fails with exit status 1 but it seems that the exit status is not passed to the docker process itself because Bamboo flags the build task as success.

30-Sep-2019 15:43:53    npm ERR! Exit status 1
30-Sep-2019 15:43:53    npm ERR!
30-Sep-2019 15:43:53    npm ERR! This is probably not a problem with npm. There is likely additional logging output above.
30-Sep-2019 15:43:53    
30-Sep-2019 15:43:53    npm ERR! A complete log of this run can be found in:
30-Sep-2019 15:43:53    npm ERR!     /root/.npm/_logs/2019-09-30T13_43_53_321Z-debug.log
30-Sep-2019 15:43:54    Stopping containers 
30-Sep-2019 15:43:54    b3efbe0710cc
30-Sep-2019 15:43:54    Removing containers
30-Sep-2019 15:43:55    b3efbe0710cc
30-Sep-2019 15:43:55    Finished task 'Script' with result: Success

The container is started with the docker run command.

Any advice?
Thanks in advance.
Regards

Setting up an SSH-tunnel using plink

cambiph — Mon, 01 Jul 2019 07:55:39 +0000

At our company, we can only connect to the outside world using the standard ports. Outgoing traffic to port 8081 is blocked by the firewall. This is a real pain in the ass. We need to publish our Javascript libraries to a Nexus instance. This instance is running in the cloud on, you guessed it, port 8081.

We are already using Jenkins pipelines to automate our workflow. The decision was made to add a new step to the pipeline: ssh-tunneling. Our Jenkins instance is running in Windows Server 2012 R2. Using bash to create a tunnel cannot be achieved. We took a look at Powershell to set up the SSH tunnel but it seems thatthe implementation of SSH in Powershell is not mature enough. Batch scripts to the rescue!

Most of us are using Putty to handle our SSH-sessions. The people who make Putty also make plink, a command-line interface to the Putty back-end. In combination with batch files we can create an SSH-tunnel like this:

plink -v -x -a -T -C -noagent -ssh -pw “password” -L 8081:our-nexus-instance.com:8081 username@our-nexus-instance.com

Let me explain the different flags used in the command above:

-v : show verbose messages
-x: enable X11 forwarding
-a: enable agent forwarding
-T: disable pty allocation
-C: enable compression
-noagent: disable use of Pageant (only needed when using SSH-keys)
-ssh: use the SSH-protocol
-pw: provide the password
-L: forward local port to remote address

Using the -L flag we can map a local port to a remote port. This is done using the following syntax: ::. This is followed by the hostname we want to connect to, prefixed with the username, used to connect: username@remoteHost.

By inserting this command into our Jenkinsfile, we can open a tunnel to our remote Nexus instance:

stage(‘Publish’) {
 steps {
  bat “plink -v -x -a -T -C -noagent -ssh -pw ‘password -L 8081:our-  
       nexus-instance:8081 username@our-nexus-instance.com”
   }
 }

Working with private repos in Docker

cambiph — Fri, 31 May 2019 09:39:45 +0000

Hello,

I am trying to work out an elegant way to build Node apps in Docker containers which use a private repo for its dependencies (JFrog artifactory) and a private Github repo.

Artifactory

I could just copy in my .npmrc but I am quite hessitant because it contains my authentication key.

Github repo

I could copy in my SSH-key but that seems like a big security risk.

What are your thoughts on this?
Thanks in advance!
Cheers