Bash script gets task log error

Octopus Server
, ,
1

Hey there,

Got my 2020.1 Octopus up and running and testing some Bash scripts aginst one of my Linux servers, but keep getting a strange result.

The following command gives an error 1 and not 0 as it should.

smbclient //server/share Password -W server -U user -c "get serverfile localfile"

it works just fine when logged into the server and switching to root, as the tentacle is running as.

I can see this i my log

Process /bin/bash in /etc/octopus/USER/Work/202011111111-4962-104 exited with code 0

Best regards
Thomas

2

Hi @thomas3

Thanks for reaching out - this is certainly an odd one. One of the first places I would look here is at the script that gets called by Octopus itself. By default, this is deleted when the deployment fails, but you can save a copy of this directory by using these steps: https://octopus.com/docs/support/copy-working-directory

Hopefully this gives you some insight as to why it’s failing!

4

Hi Justin,

Thanks for the tips, but they didn’t really give that much information back. The code is published to my server as it should. But every copy command gives an error 1 in Octopus, even if it doesn’t on the server… That is very strange.
I have now also tired to use smbget and that has the same issue.
smbget smb://user:password@server/share/file

This is what I see in my Octopus log
[file] 0b of 1.88kB (0.00%) at 0b/s ETA: Unknown April 18th 2020 08:29:15 Error

So it does copy the file over after the file copy I put in an “echo $?” and this returns a 0, so where is Octopus getting the “error 1” from!?

5

Hi @thomas3!

We typically see this in situations where console apps write their progress or other in-app messages to stderr instead of stdout (we see it a lot with things like Terraform), if you redirect your output with 2>&1 at the end of your command, it shouldn’t be seen as an error in Octopus.

Hope this helps!