Docker Community Forums

Share and learn in the Docker community.

ERR_CONNECTION_REFUSED when trying to open my software off of localhost, despite containers successfully running

I’m new to Docker, and have been reconstructing a test environment with instructions that were left to me by the previous developer.

We have several in-house pieces of job tracking software that both the outside clients and internal employees can access.

The job tracking software has a few dependencies that require a mailhog container to be spun up, and a second container which contains all of the other required sub software.

Despite my amateur knowledge on the subject, and the complexity of the software, I successfully installed Docker with the required Linux Kernels and Ubuntu.
I pulled the required images I needed, and successfully built the other ones.
I configured the in-house software with all of the testing settings enabled, and have it all grouped in the proper file directory.

By all intents and purposes, it should be configured fine.
The trouble starts with starting the containers.

The command line prompt “docker-compose up -d” pointed in the proper sub-directory throws this error:
Error response from daemon: Get “”: remote error: tls: handshake failure

I figured screw it, I’ll build it right from the Docker Desktop instead of using command line, and it worked! The two containers I needed successfully built.

I even ran the docker test command to make sure they were really running.

So here’s where the trouble just gets worse.

The ex-developer states the jobtracking software can then be viewed in browser using this “https://localhost/jobtracker/

The page doesn’t load, and then throws err_connection_refused error.

I am at my wits end with trouble shooting, because I quite simply don’t have the networking or development knowledge to search the right things, AND to make matters even more frustrating is the remote outsourced devs have their environment up and running, despite using the exact same instructions I have.

So I’m down to two possible issues. Either somewhere in the set-up process, I messed something up, OR my office network security is breaking something. However, I was told by out lead IT specialist, I should have all of the same network permissions the outsourced guys on the vpns have.

I am here now reaching out on several docker web sources to find a solution.

I am on WIndows 10.

Thankfully my employers are pretty cool about me learning as I go, and understand the technical difficulties. I just don’t want to squander the opportunity, and would like to make some progress.

How did you installed that Ubuntu. If you are on Windows, was it in the Windows Subsystem For Linux, a Virtual machine or a physical machine?

Do others use Docker Desktop or a Linux system?
When you enter inside a container can you access the jobtracker for example using curl with the IP address of the container instead of a hostname? You could also try it from the jobtracker container. If curl does not exist in the container, try wget. It wget does not exist either, install one of them.

It was on a physical machine using docker desktop.

It seems like the core root of my issues is this error:
Error response from daemon: Get “”: remote error: tls: handshake failure

A user on another forum suggested its some sort of broken certificates, but I have no idea where to start to resolve that.

I know the ex dev set this in 2019, so his version of docker is outdated by now. I think it would be ideal to get everything running properly on the current version of docker.

Actually I wanted to know more about the Linux environment since you had different error there. Sorry if I was misunderstandable. I thought you had to have another machine or a virtual machine on Windows.

Oh, sorry about the miscommunication. We are running a Linux based Ubuntu VM on our WIndows machines.


For context, I have been spending this week getting up a Docker test environment up for our inhouse job tracking software. I have been following the instructions left by the previous developer, but have been running into issues when trying to compose my containers. The first error was an TSL Handshake error that was resolved when I was whitelisted on our corporate WSA. After the WSA issue was resolved, my error has changed to this.

C:\Users\bjp\Documents\Work\universal>docker-compose up -d
[+] Running 0/1
 - universal Error                                                                                                                                                                2.0s
Error response from daemon: pull access denied for universal, repository does not exist or may require 'docker login': denied: requested access to the resource is denied

For more context, the images were built, the tracker software was properly cloned into the correct directory, the software has its own config files set for testing, and there is the yaml docker-compose file right in the root universal file directory.

I’ve been trouble shooting this for days thanks to help of folks on here, and several other forums. I’m feeling really close here. I could use the assistance to get across the finish line, so I can start developing.


Thanks for the details. Check your docker-compose.yml. It looks like it refers to an image as “universal” which does not exist and was not built before trying to use it