How can we make Linux Deploy Logs cleaner?

Hello!

Our company has some downtime to fix deployments and make our lives easier. One of the things I’d like to do is cleanup deployment logs for our Linux deployments. We get a lot of text garbage as part of the output, things like:

e[8Ae[2K
e[8Be[4Ae[2K
e[12Be[13Ae[2K

etc, usually at the beginning of lines. It looks like something is being interpreted funny by Octopus, and spitting back what should be a special character of some sort. How can we clean this garbage up, if it’s even possible?

Thanks!
Jeff

Hi @jeffrey.mitchell,

That looks a lot like ANSI color escape codes from the output of what you’re executing as part of the deployment process, which is getting fed back into the logs as part of the raw string. I’ll have a chat with our team about this, and see if we can limit the logs to printable characters only. One workaround for now could be adding a export TERM=xterm-mono to the start of your deployment script to force it into a mono terminal mode, which should strip that information out.

Hope this helps!

1 Like

It sure does. It’s a lot of docker stuff, starting containers and such. Our teams are apparently aware of this, but haven’t fixed it for every deploy involving datahub restarts. They’ve already discovered a “hide ANSI characters” flag, but didn’t apply it to the projects I primarily work on. Thanks!

2 Likes