Nextflow: command not found

Training
1

Hi Nextflow Team, this might be a dumb question: when I run nextflow run hello-world.nf under /workspaces/training/hello-nextflow directory, I get bash: nextflow: command not found. I wonder why this is happening? Appreciate your help

2

Hi @MiaZ20, welcome!

Could you provide a bit more information about how you’re doing the training? Are you using GitHub Codespaces?

Phil

3

Hi Phil @ewels ,
Thanks for your response. Yes, i’m using Github Codespace.
I’m also attaching a screenshot for your reference.
image

4

Hi @MiaZ20,

I just had a go myself, I think I know what’s happened. You will have created a GitHub Codespaces environment by following the link in the training, which boots up a space pinned to version 2.1.1 of the repository. That had a bug in the codespace configuration which caused it to boot up in “recovery mode”. That omits all of our setup, including the installation of Nextflow, which is why it’s not present.

You can see more in this GitHub issue:

The good news is that this is already fixed. If you create a codespace from the master branch today it should work fine (see settings, or go from the GitHub repository).

We will release a new version of the training today so the link on training.nextflow.io will soon work again (probably version 2.1.2).

Thanks!

Phil

5

Hi Phil, branching from master fixed this. Thank you for the help

1 Like
6

Hello,

I just created a github codespace from the master branch. I’m still getting the same error where the codespace runs in recovery mode.

7

Hi @Kterpis,

Yes you were unlucky - there was another, different error. It seems that the environment that we had created was quite close to the maximum disk space available to codespaces, and at the start of the week some unknown upstream change nudged it over that limit. We saw the same thing that suddenly new codespaces were throwing an error mid-week.

In response, I have rewritten the devcontainers setup and the docker image / environment is now much smaller. This launches without any problems. We’re about to push a new release of the training with these updates (though I think that even the older branches should now be working again, as they use the :latest tag of the base image).

TL;DR; Yes it was briefly broken for a couple of days this week, should be fixed again now. Apologies for the inconvenience as we work our way through these teething issues :folded_hands:

Phil