Conversation
… in reserved range
Disable passthrough option and make it default ON
| action='store_true', | ||
| help=('Exit with the salt call retcode and not the salt binary ' | ||
| 'retcode.') | ||
| ) |
There was a problem hiding this comment.
I am against removing this, until it goes through a deprecation path, because there are tools out there that use --retcode-passthrough, and they will need to be updated.
There was a problem hiding this comment.
@gtmanfred ah you're right, backward-compat. I will add a stub with complaint here so the CLI won't fail.
|
@isbm This test failure is related: https://jenkins.saltstack.com/job/PR/job/salt-pr-linode-ubuntu14-n/24184/ There's a small lint error too: https://jenkins.saltstack.com/job/PR/job/salt-pr-lint-n/23139/violations/ :) |
|
@rallytime hopefully now is fine. But I touched another test, let see... :) |
|
👌 Tests look good to me. |
terminalmage
left a comment
There was a problem hiding this comment.
This is a pretty substantial change and needs some input from @thatch45. The major changes include changing the return code range of the various salt-ssh steps, as well as defaulting --retcode-passthrough to being always on.
| self.add_option('--retcode-passthrough', | ||
| default=False, action='store_true', | ||
| help=('Exit with the salt call retcode and not the salt binary return code. ' | ||
| 'NOTE: this is now always ON and is here only for backward compatibility.')) |
There was a problem hiding this comment.
The help message seems to conflict with the default value here.
terminalmage
left a comment
There was a problem hiding this comment.
We've already decided not to default --retcode-passthrough to being enabled in #48361. Also, I'm not really sure that such drastic changes are warranted to salt.defaults.exitcodes.
|
As I understand it, these changes were superceded by the changes made recently by @terminalmage. Is that correct and if so shall we go ahead and close this? |
What does this PR do?
Error codes cleanup
Previous Behavior
Error codes are hard-coded, brings meaningless data, overrides standards
New Behavior
Organised.
Tests written?
No
@terminalmage this helps the #48361