[22:37] <ctracey> hey guys, ltns
[22:37] <ctracey> i think i will have 2 small changes to push shortly
[22:38] <ctracey> before i push the bigger of the 2, just curious if folks ever have issues with package_command from within cloud-init
[22:38] <ctracey> from time to time on RHEL we had issues with yum just failing...no clear reason why
[22:53] <harlowja> hmmmm, yum is just problematic in general
[22:53] <harlowja> *imho its a pile of crap, ha
[22:55] <harlowja> what errors do u see ctracey 
[22:55] <ctracey> hey, i wouldnt disagree
[22:55] <ctracey> well, we dont see errors only bc capture=False
[22:55] <harlowja> hmmm
[22:56] <ctracey> so, my change is simply to allow for debugging package_command
[22:56] <ctracey> just wondering if anyone else had encountered it
[22:56] <harlowja> i haven't
[22:57] <harlowja> but i haven't used that feature to much
[22:57] <ctracey> basically I would propose system_info: debug_package_command: true
[22:57] <harlowja> seems to make sense
[22:57] <ctracey> to you can at least grab the output as necessary
[22:57] <harlowja> yahoo has a whole other way to install stuff (like juju in a way)
[22:57] <ctracey> ah, interesting
[22:57] <harlowja> but ya, i think debug package command makes sense
[22:57] <ctracey> the other issue we ran into was a 2.7 vs 2.6 thing
[22:58] <ctracey> buried in the AzureDataSource
[22:58] <harlowja> ah, blame smoser for that one, ha
[22:58] <harlowja> :-P
[22:58] <ctracey> (yes, we are on an ancient python release)
[22:58] <ctracey> :)
[22:58] <ctracey> its a simple fix
[22:58] <ctracey> ill get them in tonight most likely
[22:58] <harlowja> cool
[22:58] <harlowja> sounds great
[22:59] <ctracey> harlowja: you going to HK?
[22:59] <harlowja> sure am
[22:59] <ctracey> niiiice
[22:59] <ctracey> i'll see ya there then :)
[23:00] <harlowja> cool
[23:59] <smoser> ctracey, yeah. i'm sorry about that.
[23:59] <smoser> but wrt debug_package_command
[23:59] <smoser> that output shoudl be caught by 
[23:59] <smoser> output: {all: '| tee -a /var/log/cloud-init-output.log'}
[23:59] <smoser> is it not?