| jtv | This is annoying. I'm trying to Q/A my uec2roottar rewrite, but something else is broken. | 03:38 |
|---|---|---|
| jtv | Could not find kernel image: amd64/generic/trusty/release/boot-kernel | 03:38 |
| jtv | I think it's an architecture mismatch: i386/amd64 | 03:39 |
| jtv | Silly of me. Can't set a node's architecture with default enlistment. | 03:41 |
| === vladk|offline is now known as vladk | ||
| === vladk is now known as vladk|offline | ||
| === CyberJacob|Away is now known as CyberJacob | ||
| === vladk|offline is now known as vladk | ||
| === vladk is now known as vladk|offline | ||
| jtv | rvba: might not be completely crazy to start a DownloadProgressHandler. | 07:32 |
| rvba | jtv: that's what I thought. The thing is that the documentation for this API method is so detailed that it drives people to try and use that method :) | 07:32 |
| jtv | On the other hand, the design notes had to go _somewhere_. :) | 07:32 |
| rvba | heh | 07:33 |
| jtv | If we just provide a GET to download progress, that wouldn't be very user-friendly but at least API users could get stuff done. | 07:33 |
| rvba | Indeed. | 07:33 |
| jtv | And it would get the implementation ball rolling: so much easier to say how it should be _better_ than how it should _be_. :) | 07:33 |
| rvba | Also, isn't the import script supposed to do the reporting (using the report_download_progress method)? | 07:34 |
| rvba | Wasn't that the plan? | 07:34 |
| === CyberJacob is now known as CyberJacob|Away | ||
| rvba | jtv: something else I wanted your opinion about: When users upgrade from Saucy, the new entries in bootresources.yaml don't have a 'label' constraint. As a result, multiple images can be imported (for instance 'rc'/'release')… shouldn't we add a labels=['release'] constraint? This would be better in sync with what the old config (from which we're migrating) was doing. | 08:07 |
| rvba | Not sure if it's worth a bug/fixing. Especially considering the upcoming work on getting rid of bootresources.yaml. | 08:07 |
| jtv | rvba: maybe wait until we know what we really want, there? Otherwise we keep modifying and creating this massive history of different default settings. | 08:25 |
| rvba | jtv: yeah. But let's keep that problem in mind then. | 08:25 |
| jtv | Absolutely. | 08:25 |
| jtv | gmb, this one's for you: https://code.launchpad.net/~jtv/maas/test-1310844/+merge/217006 | 09:23 |
| gmb | jtv: On it, ta | 09:27 |
| rvba | bigjools: can you please vote on https://code.launchpad.net/~rvb/maas/add-power-params-doc/+merge/216894 ? | 09:31 |
| rvba | jtv: reviewing your 'python-to-python-imports' branch now | 09:36 |
| rvba | allenap: I filed https://bugs.launchpad.net/maas/+bug/1312085 | 09:44 |
| ubot5 | Launchpad bug 1312085 in MAAS "No documentation for the CLI" [High,Triaged] | 09:44 |
| === vladk|offline is now known as vladk | ||
| rvba | allenap: care to vote on https://code.launchpad.net/~rvb/maas/add-power-params-doc/+merge/216894 ? | 11:31 |
| allenap | rvba: Okay, I’m finishing something off right now, then I’ll do it. | 11:37 |
| rvba | allenap: ta, it's definitely not urgent. | 11:41 |
| alfs | ran into an interesting issue with usb keys and large rams. | 12:42 |
| alfs | The machine has 24 GB ram, and I use a 8 GB usb stick for deployment. | 12:43 |
| alfs | the rootfs gets about 1 GB, and the rest is allocated for swapspace - and e.g. juju bootstrap fails due to insufficient disk space. | 12:44 |
| alfs | is there some easy way to skip swap altogether, or do I need a customized preseed? | 12:47 |
| Teduardo | if using MaaS to build/maintain a distributed system, lets use for example openstack. when a new openstack release comes out, can MaaS sensibly handle the upgrade as well? | 12:52 |
| Teduardo | (maas/juju) | 12:53 |
| magicrobotmonkey | I'm getting PXE-E11 ARP Timeout when booting a potential node | 13:07 |
| melmoth | what is the purpose of the new "Network" menu in trusty maas ? | 14:04 |
| melmoth | i did not have such a one before using the ppa version of maas in precise. what am i suppose to put there ? | 14:04 |
| rvba | melmoth: maas.ubuntu.com/docs1.5/networks.html | 14:06 |
| melmoth | ok, so if all my nodes are on the same net as the nic i assigned to the cluster controller , and dont use other network, i dont need to use this form, right ? | 14:07 |
| rvba | melmoth: correct | 14:08 |
| melmoth | good. thanks. | 14:08 |
| magicrobotmonkey | my node came to the maas-enlisting-node login: prompt | 14:08 |
| magicrobotmonkey | am I supposed to do something now? | 14:08 |
| rvba | magicrobotmonkey: no, this should be temporary (i.e. until your node enlists itself). After a while your node should be powered down by MAAS. | 14:09 |
| magicrobotmonkey | oh there it goes | 14:10 |
| magicrobotmonkey | things are happening! | 14:10 |
| magicrobotmonkey | cool it showed up in the nodes | 14:10 |
| magicrobotmonkey | this is much better than scraping the switch for mac addresses | 14:11 |
| magicrobotmonkey | when i hit "commision node" where should I look for a log of whats going on? | 14:19 |
| === vladk is now known as vladk|offline | ||
| alfs | console output | 14:53 |
| alfs | I find is the best. Sometimes the commissioning or running hangs, and the console is the best place to see this | 14:54 |
| magicrobotmonkey | yea ok | 14:54 |
| magicrobotmonkey | does the commisioning configure ipmi locally? because I've been unable to use ipmitool on these boxes but maas did it somehow | 14:55 |
| alfs | It adds it's own user | 15:06 |
| alfs | in the commissioning stage | 15:06 |
| magicrobotmonkey | ah cool | 15:06 |
| magicrobotmonkey | thats pretty slick | 15:06 |
| alfs | still no luck with partitioning. Is there some other source for the partition layout than /etc/maas/preseeds/preseed_master? | 15:13 |
| alfs | I know my edits are working, e.g. changing from ext4 to ext3, but partman-auto/expert_recipe and partman-auto/choose_recipe just does not bite. | 15:14 |
| rvba | alfs: that's weird, there is nothing that should override changes in /etc/maas/preseeds/preseed_master. | 15:15 |
| alfs | Hm.. I commented everything regarding partman in the preseed, and then it (as intended) stops at the partitioning stage | 15:35 |
| alfs | However I cannot get past iscsi config, i.e. I choose iscsi, finish, and then get thrown back to iscsi config. | 15:35 |
| alfs | No option to configure, and if I finish partitioning it (of course) has no root fs defined | 15:36 |
| alfs | perhaps due to the partman early-command not being run to define the disks.. enabling that and trying again | 15:37 |
| alfs | seems there was a small typo in the expert_recipe, causing it to be ignored, seems to be working now. | 16:13 |
| magicrobotmonkey | heh yea i think i had a problem like that at one point | 16:17 |
| === roadmr is now known as roadmr_afk | ||
| === roadmr_afk is now known as roadmr | ||
| magicrobotmonkey | do maas nodes use a proxy by default? | 17:51 |
| magicrobotmonkey | an apt proxy, that is | 17:51 |
| === CyberJacob|Away is now known as CyberJacob | ||
| Isid0r0 | Hey, Does anyone know how to add the new "trusty" disto to an existing MAAS installation? I see it downloaded the OS's but it isn't registering in the settings page as a deployable distro | 18:11 |
| === roadmr_ is now known as roadmr | ||
| === kevin is now known as Guest48000 | ||
| === kev is now known as Guest49483 | ||
| alfs | magicrobotmonkey: Yes, the maas server has squid-deb-proxy installed | 19:39 |
| magicrobotmonkey | cool | 19:39 |
| magicrobotmonkey | thats super handy | 19:39 |
| alfs | make sure you use the official archive, otherwise you get bad mirror hangs | 19:40 |
| alfs | or {cc}.archive.ubuntu.com at least | 19:40 |
| === CyberJacob is now known as CyberJacob|Away | ||
| === CyberJacob|Away is now known as CyberJacob | ||
| === CyberJacob is now known as CyberJacob|Away | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!