[02:45] <timClicks> is any special syntax required for cross-controller and/or cross-cloud CMR?
[03:17] <hpidcock> wallyworld: did your recent fixes fix this https://bugs.launchpad.net/juju/+bug/1880422 ?
[03:17] <mup> Bug #1880422: Bootstrapping manual controller on arm64 fails <manual-provider> <juju:Triaged> <https://launchpad.net/bugs/1880422>
[03:17] <wallyworld> hmmm
[03:18] <wallyworld> maybe not
[03:18] <wallyworld> i'd need to look at the code
[03:19] <hpidcock> I can take a look
[03:19] <wallyworld> i think he just needs ---constraints "arch=arm64"
[03:19] <wallyworld> which is separate to my fix, that should already have worked
[04:05] <wallyworld> thumper: https://github.com/juju/juju/pull/11731
[04:39]  * thumper looks
[04:58] <thumper> wallyworld: done
[04:59] <wallyworld> ty
[05:06]  * thumper sighs...
[05:06] <thumper> I though I had all this sorted, now got weirdness
[05:10]  * thumper takes a deep breath
[05:10] <thumper> figured out why some of them are failing
[05:10] <thumper> FFS
[05:14] <hpidcock> thumper: did you turn it off and on again?
[05:20] <thumper> hpidcock: oh I wish it was that easy
[05:21] <thumper> now to go and fix tests that I've broken with the "fix"
[05:21] <thumper> this is the branch that will never die
[05:28] <thumper> ok... I think they are all passing now
[05:28] <thumper> if that's right, it will now be time to pull the branch apart to land in pieces
[05:28] <thumper> ugh... still need to write the upgrade step
[05:28] <thumper> can do that later
[05:28] <thumper> won't be landing that bit first anyway
[05:34]  * thumper pulls off a chunk to land
[05:46] <thumper> wallyworld, hpidcock: https://github.com/juju/juju/pull/11732
[05:46] <hpidcock> lookin
[05:47] <thumper> simple PR to just move stuff
[05:48] <thumper> hpidcock: I'm EODing, if you're happy, can you please $$merge$$ it?
[05:48]  * thumper out
[05:48] <hpidcock> sure thing
[08:13] <achilleasa> flxfoo: curious why you got the usage prompt; perhaps adding 'set -eux' at the top of the function should provide more info about what gets passed to nc? If you do come up with an improved version of the script by all means post it to that page for others to use
[10:05] <manadart_> achilleasa: Got time to review this one? Follows the original refactor you reviewed: https://github.com/juju/juju/pull/11734
[10:07] <achilleasa> manadart_: sure
[10:09] <flxfoo> achilleasa:I think the ssh client version with proxycommand change, not to use nc anymore
[10:30] <stickupkid> manadart_, achilleasa CR https://github.com/juju/juju/pull/11733
[11:54] <hml> good morning
[12:47] <manadart_> Morning hml.
[12:47] <manadart_> Anyone able to review a forward merge? No conflicts. https://github.com/juju/juju/pull/11735
[14:09] <hml> stickupkid: review please https://github.com/juju/juju/pull/11722
[14:16] <stickupkid> hml, nice :+1:
[14:20] <hml> stickupkid:  ty
[14:21] <rick_h> petevg:  do you have time to pre-forrester?
[14:21] <rick_h> pre-wave?
[14:27] <manadart_> stickupkid: https://github.com/juju/description/pull/79
[14:28] <petevg> rick_h: I've got TGIF at 11am, my time. I'm free now, though. Want to jump into a hangout?
[14:29] <rick_h> petevg:  sure, tell you what let's hop in your tgif and I'll just eat the gap if that's ok
[14:29] <petevg> Sounds good :-)
[14:31] <rick_h> guild heads up I'm going to steal guimaas for demo work again kthx sorry
[14:31] <stickupkid> achilleasa, ^
[14:32] <achilleasa> stickupkid: got my own maas now :p
[14:44] <rick_h> achilleasa:  ooooh, fancy!
[14:52] <achilleasa> rick_h: there wasn't an easier/quicker way to test the OVS stuff so I finally bit the bullet and set up a local with the edge snap :D
[14:54] <rick_h> achilleasa:  cool
[15:53] <stickupkid> hml, find https://github.com/juju/juju/pull/11736
[15:54] <hml> stickupkid: osm is a bundle, so i’m guessing the bundle contains wordpress
[15:54] <hml> stickupkid: so that makes sense to me for find
[15:54] <stickupkid> hml, osm is the k8s stuff
[15:55] <stickupkid> hml, https://jaas.ai/osm/bundle/44
[15:55] <hml> stickupkid: maybe not… i’ts a bundle, but no wordpress
[15:55] <stickupkid> hml, yeah... interesting though
[15:57] <stickupkid> hml, also it states that it's a charm, when it's a bundle, so we really should catch that?
[15:57] <hml> stickupkid: bug in api?
[15:57] <hml> stickupkid: we do catch it… just need to type the Type :-)
[15:58] <stickupkid> hml, it's really intesting, that they have a leaf call "charm"... I wonder if there is a better word for that
[15:59] <stickupkid> i.e. we have "type: bundle" and a leaf called "charm"
[15:59] <stickupkid> you get me?
[16:02] <hml> stickupkid: not following
[16:02] <stickupkid> ho?
[16:02] <hml> stickupkid: sure
[16:14] <rick_h> petevg:  did get the bootstrap to work with the new larger instances
[16:19] <petevg> rick_h: cool. I'm still working on azure (I've got my credentials all straightened out. I'll see what I can do to make sure that I get instances that are large enough myself.)
[16:36] <petevg> rick_h: how much ram and disk did you end up allocating to each node?
[16:38] <rick_h> petevg:  8gb ram k8s nodes
[16:38] <rick_h> I think they're two cores
[16:38] <rick_h> I think the key was getting from 4gb of ram default notes to the 8gb of ram ones
[16:41] <petevg> rich_h: cool. Thx!
[17:36] <thedac> coreycb: https://pastebin.canonical.com/p/KvKPxrMTkK/
[17:36] <thedac> cat /root/.pydistutils.cfg
[17:36] <thedac> [easy_install]
[17:36] <thedac> find_links = file:///var/lib/juju/agents/unit-octavia-0/charm/wheelhouse/
[17:36] <thedac> allow_hosts = ''
[17:36] <thedac> pip list |grep setup
[17:36] <thedac> setuptools     47.3.1
[17:36] <thedac> setuptools-scm 1.17.0
[17:36] <thedac> dpkg -l |grep setuptoo
[17:36] <thedac> ii  python3-setuptools               39.0.1-2                                    all          Python3 Distutils Enhancements
[17:36] <thedac> Could it be checking the wrong version of setuptools?
=== psydroid is now known as Guest84896
[17:42] <thedac> coreycb: setuptools_ver = _load_installed_versions('pip3').get('setuptools') is returning None
[18:46] <petevg> rick_h: what demo are you deploying on your k8s cluster? Just mongodb, or something else?
[18:49] <rick_h> petevg:  I'm getting a small kubeflow setup from the k8s team
[18:50] <petevg> rick_h: nice. Share it with me when you get it? (I've got a model theoretically running on AKS, but I need to test it to verify.)
[18:51] <rick_h> petevg:  yea, I'm using hte mongodb one to test with. Let me get you the bundle I'm using
[18:52] <rick_h> petevg:  https://pastebin.canonical.com/p/nrdyYwXMhw/
[18:52] <petevg> thx!
[18:54] <rick_h> petevg:  yea so that's just mongodb.yml and juju deploy ./mongodb.yaml and should work
[18:54]  * rick_h has to run the boy to a party, I'll check back in later
[19:41] <petevg> rick_h: AKS was pretty straightforward to bootstrap. Where would you like me to drop my notes?
[20:51] <rick_h> petevg:  email would be fine for now if that's cool
[21:20] <petevg> rick_h: sent. Have a great weekend!
[21:35] <rick_h> petevg:  ty you too