Simplify acceptance, allow hack for Debian 8 to return 1 or 100 #413
Conversation
wyardley
commented
Oct 24, 2017
wyardley
commented
- simplify acceptance tests
- allow the hack to install openhpid on Debian 8 to continue even if it exits 1 or 100 (100 seems to be what the error is making it exit).
| # For Debian 8 "jessie", we need | ||
| # - pacemaker and crmsh delivered in jessie-backports only | ||
| # - openhpid post-install may fail (https://bugs.debian.org/785287) | ||
| if fact('lsbdistcodename') == 'jessie' | ||
| if fact('os.family') == 'Debian' && fact('os.release.major') == '8' |
There was a problem hiding this comment.
I liked your proposal to use the codename fact to simplify the conditional in #411, it makes the code simpler.
There was a problem hiding this comment.
Actually, I had suggested it mostly if we were doing it for both 7 / 8. I could switch it back, but I think maybe this is clearer (and also forget whether the lsb* facts are as reliably there?)
There was a problem hiding this comment.
I like the os.release.major approach.
| on host, 'echo deb http://ftp.debian.org/debian jessie-backports main >> /etc/apt/sources.list' | ||
| on host, 'apt-get update && apt-get install -y openhpid' | ||
| on host, 'apt-get update && apt-get install -y openhpid', acceptable_exit_codes: [0, 1, 100] |
There was a problem hiding this comment.
Indeed, apt-get rc is 100 when a sub-command fails.
| # For Debian 8 "jessie", we need | ||
| # - pacemaker and crmsh delivered in jessie-backports only | ||
| # - openhpid post-install may fail (https://bugs.debian.org/785287) | ||
| if fact('lsbdistcodename') == 'jessie' | ||
| if fact('os.family') == 'Debian' && fact('os.release.major') == '8' |
| # For Debian 8 "jessie", we need | ||
| # - pacemaker and crmsh delivered in jessie-backports only | ||
| # - openhpid post-install may fail (https://bugs.debian.org/785287) | ||
| if fact('lsbdistcodename') == 'jessie' | ||
| if fact('os.family') == 'Debian' && fact('os.release.major') == '8' |
There was a problem hiding this comment.
I like the os.release.major approach.
|
now tests are failing for centos 7 too |
|
The acceptance test for centos 7 will fail until we finalize and merge #410 . |
|
Then thank you both :) |
