New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Highlight recent git tags (e.g. "8.0-preview-2" and "v0.5.5.9") in /usr/bin/iiab-summary, iiab-diagnostics etc [and remote URLs, branch names, PR count / commit count since tag...] #3267
Conversation
is repo anywhere? repo, branch, tag, commit is a pretty strong identifier. git remote -v | grep '(fetch)' (extra marks for awking off the (fetch) part) |
Yes. Both remotes are near the top of iiab-diagnostics output, and also /usr/bin/iiab-summary output, e.g. here is sample RPi output:
|
Here is sample /usr/bin/iiab-summary output on Ubuntu:
|
iiab-diagnostics (and other community tools) should benefit from — and co-evolve with — several aspects of the very concise |
Given iiab-apps-to-be-installed is now a subset of iiab-diagnostics would it be better to install these new files at the same time as iiab-diagnostics? Plays nicer with a git pull without needing to repeat 1-prep given the future code churn going forward. |
I thought about that. Definitely a good idea to ponder but also needs some thinking, as we strive to make Stage 0 (roles/0-init) as lightweight + quick as possible. Using var |
Do I really have to file a bug about
Should of thought about that harder when 21f0857 was pushed. I'd rather see a symlink used as /usr/bin/filename pointing to /opt/iiab/iiab/scripts/filename. Still would need a pass though ansible once to put the symlinks in place but a pass though ansible would not be required when the scripts are updated later with a git pull as the symlink are always pointing at the git's current code. Could just use the full path /opt/iiab/iiab/scripts/iiab-apps-to-be-installed within iiab-diagnostics as an in between compromise. |
Yes, this is fully expected: iiab-diagnostics is designed to be resilient — when calling fragile commands that fail in all sorts of ways (within cat_cmd).
It's an option. In the end I'm hoping roles/0-init can be sped up this year, very dramatically if possible. To speed up common testing paths and make them much more understandable. No doubt there are different ways to make these things happen — so long as it happens, this year hopefully (even if none of our short-term options are beautiful, no doubt ;) |
Perhaps some elements of #2961 might be useful
note check-vars utility, with iiab-from-console.yml falling out of favor think a move to a 'pre-check' within iiab-install away from 0-init would cut the runtime within admin-console. Given admin-console wrote out any changes if needed I'd be comfortable with assuming they are correct and not waste the time to validate.
Food for thought. |
Good ideas. My quick profiling of roles/0-init earlier today showed validate_vars.yml generally completes in under 5 seconds, and is very much worth it — but as you suggest such things can be outsourced to bash/Python/C/whatever for further speed-ups in future if people find that necessary. |
Those same 5 seconds are wasted given runrole deals with a single role as the validation is covering every role, that is one of the gains with moving validate to it's own role in 2961 |
Uses
git describe --tags --abbrev=0
to record "most recent" Git tag(s) like "8.0-preview-2" and "v0.5.5.9" to various places — e.g. /etc/iiab/iiab.ini and iiab-diagnostics pastebin.Ref: