| tombee | Hey, does anyone have any examples on how to set up LVM for mounted disks (non-root partitions) with cloud-init? | 12:32 |
|---|---|---|
| rharper | https://canonical.my.salesforce.com/5003z00001yXYRIAA4 | 15:51 |
| rharper | Odd_Bloke: so for the Oracle Datasource issue; I believe this is what they see _on_ OracleLinux using updated cloud-init datasource | 15:51 |
| rharper | hence the "dracut" initramfs fix at the bottom | 15:51 |
| Odd_Bloke | Right, I think I understand the issue there, based on a quick scan through. | 15:54 |
| AnhVoMSFT | rharper any concern in landing this https://code.launchpad.net/~vtqanh/cloud-init/+git/cloud-init/+merge/369785 ? I looked at current analyze code and it looks like it's only finding events that are of type start: or finish:, so these events won't be even touched by analyze (which is what we want) | 17:06 |
| AnhVoMSFT | for stacking azure-ds events nicely within analyze (which will be another PR), I think one possible way to handle it would be to make an educated guess of which cloud-init phases we're in when we see those events (during analyze) based on the timestamps. Otherwise we would have to refactor the code to make args.reporter global so that other reporte | 17:08 |
| AnhVoMSFT | rs can find it as the parent | 17:08 |
| rharper | AnhVoMSFT: do you have a cloud-init analyze show/blame output with those new events in place ? | 17:44 |
| AnhVoMSFT | rharper https://paste.ubuntu.com/p/mfc4YSbFQQ/ | 17:52 |
| rharper | AnhVoMSFT: right; we still have to do something for https://bugs.launchpad.net/cloud-init/+bug/1833731 | 17:55 |
| ubot5 | Ubuntu bug 1833731 in cloud-init "cloud-init analyze output not formatted cleanly on Azure" [Undecided,New] | 17:55 |
| rharper | AnhVoMSFT: but the diagnostic event doesn't contribute to the formatting issue; we should land something for the current formatting and your additional events as well | 17:56 |
| AnhVoMSFT | yep, we do need to fix that, but like you said it shouldn't be the blocker for those additional events | 17:56 |
| AnhVoMSFT | I'm trying to explore the options for fixing it | 17:56 |
| rharper | ok | 17:59 |
| samgilson | hey Odd_Bloke, on my MP you just approved the CI bot flagged my commit message for not meeting the style standards. I just fixed it, but is there any way to re-run it from my side to make sure it is correct? | 19:40 |
| rharper | samgilson: no, we can poke it again | 19:41 |
| rharper | samgilson: ok, I've remarked approved | 19:43 |
| AnhVoMSFT | rharper What do you think about this fix for analyze? https://paste.ubuntu.com/p/zHyDKCbWVr/ - it's relatively simple. WIP branch: https://code.launchpad.net/~vtqanh/cloud-init/+git/cloud-init/+ref/UpdateReporterForAnalyze | 19:43 |
| samgilson | rharper thanks so much! | 19:44 |
| AnhVoMSFT | the names are now showing up correctly, but I think we need to land your fix that would let analyze process more than 2 level deep to be able to render it correctly | 19:44 |
| AnhVoMSFT | Also I updated the merge proposal for the additional telemetry addressing your comments | 19:44 |
| rharper | thanks; | 19:45 |
| AnhVoMSFT | rharper if I apply your commit here https://paste.ubuntu.com/p/y3JskPgtg8/ on top of that UpdateReporterForAnalyze WIP branch and run cloud-init analyze show I get this nicely formatted output: https://paste.ubuntu.com/p/y3JskPgtg8/ | 20:34 |
| rharper | AnhVoMSFT: \o/ ; though I think we shoudl somehow get a DataSourceAzure prefix in those somehow | 20:36 |
| AnhVoMSFT | wrong link to the commit : https://git.launchpad.net/~raharper/cloud-init/commit/?id=fe1d854dbb63fa6c8219547473cdc02a7fb02709 | 20:36 |
| rharper | that does look much better | 20:36 |
| rharper | would like to get a init-local/datasource/_get_data which would give us one more indent level I think; | 20:37 |
| rharper | well, s/datasource/$name_of_the_Datasource; but I think it may be easy enough to set that in your event wrappers rather than some how read it from the current ds object | 20:38 |
| AnhVoMSFT | yeah, the event wrapper can probably inject azure_ds: in there | 20:39 |
| Odd_Bloke | rharper: https://code.launchpad.net/~xiaofengw/cloud-init/+git/cloud-init/+merge/368902 <-- did you accidentally kick off the autoland test job manually for this MP? | 20:44 |
| Odd_Bloke | (It's not a problem if you did, just trying to understand the error message!) | 20:44 |
| rharper | possibily | 20:44 |
| rharper | hrm; I know I added a ci run; but maybe I also submitted an autoland | 20:45 |
| rharper | the comments are off | 20:46 |
| rharper | net: skip bond interfaces in get_interfaces (detail) | 20:46 |
| rharper | I did marked that branch approved | 20:46 |
| rharper | but I don't see how it updated the vmware MP | 20:46 |
| Odd_Bloke | :/ | 20:47 |
| rharper | indeed | 20:47 |
| samgilson | hey Odd_Bloke rharper the CI bot came back around and set my MP back to needs review, the commit message should be correct now. | 21:10 |
| rharper | samgilson: thanks | 21:11 |
| Odd_Bloke | samgilson: Thanks! Convincing the bot to accept your commit message is a rite of passage for cloud-init development, so welcome to the club. ;) | 21:11 |
| samgilson | Odd_Bloke Glad to see its not just me who messes this up '=D | 21:27 |
| Odd_Bloke | samgilson: For reference, I normally do `git commit --allow-empty`, compose in $EDITOR (vim, in my case), and then `:w /tmp/commit-msg` and use that. | 21:28 |
| Odd_Bloke | (My vim is configured to wrap git commits correctly, so it solves the problem.) | 21:29 |
| samgilson | Odd_Bloke thanks! I will try it for the next one and hopefully it works =D | 21:30 |
| Odd_Bloke | Fingers crossed! | 21:30 |
| AnhVoMSFT | rharper https://code.launchpad.net/~vtqanh/cloud-init/+git/cloud-init/+merge/370156 - I "stole" your commit (cherry-picking it) into my branch on top of my changes for the MP | 21:33 |
| rharper | AnhVoMSFT: nice | 21:34 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!