#10506 - Machine Readable Global-Status

[SilverPreece]

As noted in #10506, the vagrant global-status --machine-readable command produces a rather unhelpful format for machine processing. It does not specify any @ui.machine outputs, only @ui.info, meaning that the formatted table outputted by this command is fudged into a series of machine-readable lines that are difficult to parse and not particularly useful. This PR adds @ui.machine calls alongside them, to make parsing the output of this command much easier, and consistent with similar commands such as vagrant status --machine-readable.

Before:

1559265195,,ui,info,id       
1559265195,,ui,info,name    
1559265195,,ui,info,provider   
1559265195,,ui,info,state    
1559265195,,ui,info,directory                                                        
1559265195,,ui,info,
1559265195,,ui,info,------------------------------------------------------------------------------------------------------
1559265195,,ui,info,2e5afa9  
1559265195,,ui,info,default 
1559265195,,ui,info,virtualbox 
1559265195,,ui,info,poweroff 
1559265195,,ui,info,/path/to/vagrant/file 
1559265195,,ui,info,
1559265195,,ui,info, \nThe above shows information about all known Vagrant environments\non this machine. This data is cached and may not be completely\nup-to-date (use "vagrant global-status --prune" to prune invalid\nentries). To interact with any of the machines%!(VAGRANT_COMMA) you can go to that\ndirectory and run Vagrant%!(VAGRANT_COMMA) or you can use the ID directly with\nVagrant commands from any directory. For example:\n"vagrant destroy 1a2b3c4d"

After:

1559265166,,metadata,machine-count,1
1559265166,,machine-id,62cfe71
1559265166,,provider-name,virtualbox
1559265166,,machine-home,/path/to/vagrant/file
1559265166,,state,poweroff

Please see the comments for a request for info re: suppressing UI output

I am currently writing a build system wrapper for Vagrant, so I will be using the machine-readable option on most of the available Vagrant commands. If Hashicorp are receptive to patches along similar lines to this one, I would like to help clean up the machine-readable mode and make it more consistent. It's a very useful mode to have, and makes integrating Vagrant with automation/build tools much easier.

[hashicorp-cla]

All committers have signed the CLA.

[briancain]

Hey there @GregJPreece - Thanks for taking the time to make a pull request with Vagrant!

While generally I do agree that the --machine-readable flag needs some love and attention, I don't think we should be removing the ui output as a default behavior. I can only imagine how many users at this point have built up automation around being able to read in the UI level (like for errors) and parse what vagrant prints. Removing this context completely will break those users tools, so I'm 👎 here on removing it.

One enhancement that could be worth making is having a difference between "messages" and general output, but I'm not entirely sure how to do that without adding more columns and keeping it consistent with how it works now 🤔

[SilverPreece]

Indeed, as mentioned in my comment I'm not myself in favour of breaking current default behaviour, but I'm not familiar enough with Ruby or Vagrant to know how to get a flag check into the UI lib function. Do you have any suggestions?

[chrisroberts]

Hi @GregJPreece, thanks for this PR! I am 👍 on the updates to make the output of the global-status command more useful when in machine readable mode. I don't think there's any reason to remove ui outputs (and mentioned that in the review). Otherwise, there's just a few things to be adjusted and this PR should be good to go.

Cheers!

[chrisroberts]

We would expect the UI output to still be included within the machine readable output. A new test should also be added which validates the new machine readable content is output as expected.

[chrisroberts]

Thanks for the updates. Everything looks great. I did toss in some test coverage on the new behavior, just FYI.

Cheers!

[SilverPreece]

Thanks for the updates. Everything looks great. I did toss in some test coverage on the new behavior, just FYI.

Cheers!

You beat me to it! I was going to figure those out this evening. Thanks for taking care of that for me. I may have similar patches to come, based on my experiences with machine-readable mode & automation. I'll keep the comments raised on this PR in mind for those.

[ghost]

I'm going to lock this issue because it has been closed for 30 days ⏳. This helps our maintainers find and focus on the active issues.

If you have found a problem that seems similar to this, please open a new issue and complete the issue template so we can capture all the details necessary to investigate further.