mup | Bug #1469663 opened: 1.8.0 Error validation highlight displayed to soon <ui> <ux> <MAAS:New for ricgard> <https://launchpad.net/bugs/1469663> | 09:40 |
---|---|---|
mup | Bug #1469666 opened: 1.8.0 incorrect text alignment in dropdown fields <ui> <ux> <MAAS:New for ricgard> <https://launchpad.net/bugs/1469666> | 09:55 |
mup | Bug #1469666 changed: 1.8.0 incorrect text alignment in dropdown fields <ui> <ux> <MAAS:New for ricgard> <https://launchpad.net/bugs/1469666> | 10:04 |
mup | Bug #1469666 opened: 1.8.0 incorrect text alignment in dropdown fields <ui> <ux> <MAAS:New for ricgard> <https://launchpad.net/bugs/1469666> | 10:10 |
Delfyn | the MAAS URL passed to openstack, does it need to be resolvable also by the other hosts, or only by the openstack server? I ask because I've installed MAAS and openstack on the same system, with the MAAS system specified to openstack as 127.0.0.1, but in comands.log I see: INFO juju.network network.go:194 set | 11:47 |
Delfyn | ting prefer-ipv6 to false\n2015-06-29 11:19:43 ERROR juju.cmd supercommand.go:430 Get http://127.0.0.1/MAAS/api/1.0/nodes/?agent_name=6805b4ae-d1c8-4561-8b2b-8e38072c5881&id=node-ce38fc4a-1e4 | 11:47 |
Delfyn | d-11e5-b294-e41f13cc851c&op=list: dial tcp 127.0.0.1:80: connection refused\nERROR failed to bootstrap environment: subprocess encountered error code 1 | 11:47 |
caribou | I have more questions on my Curtin customization journey | 11:53 |
caribou | I'm now able to properly partition, configure & install on a disk and it works well if this disk is the only one | 11:54 |
caribou | but if the system has /dev/vda which is the first disk & /dev/vdb which is the big disk being partitioned, | 11:55 |
caribou | how can I tell the unit to reboot on the second partitioned disk ? | 11:55 |
caribou | If I remove /dev/vda from the system & reboot the unit, it will correctly complete the charm INSTALL and the service comes online | 11:57 |
dimitern | mpontillo, hey, are you around? | 12:06 |
dimitern | rvba, ping | 12:15 |
rvba | dimitern: hey | 12:15 |
dimitern | rvba, hey | 12:15 |
dimitern | rvba, so I'm trying to follow mike's "A Portable MAAS Dev/Test Environment" presentation | 12:15 |
dimitern | rvba, and I've already mirrored the trusty archive, just the sstreams-mirror is still going | 12:16 |
dimitern | rvba, once this is done (and the additional steps in "Putting it all Together"), how do I configure my local maas to use the mirror? | 12:17 |
rvba | dimitern: configure MAAS to use this mirror by changing the "Main archive" setting (on the settings page). | 12:18 |
dimitern | rvba, can it be a path instead of an URL ? | 12:18 |
rvba | dimitern: no, it has to be a URL. The nodes themselves will use it. | 12:19 |
dimitern | rvba, so I need nginx or apache running to server the mirror? | 12:19 |
dimitern | serve | 12:19 |
rvba | Yes, you need something to serve the mirror. I don't remember what Mike used… probably a dedicated tool (I forget) the name. | 12:20 |
dimitern | rvba, right, I'll figure it out then | 12:20 |
rvba | dimitern: I'll ask Mike when he comes online. | 12:21 |
dimitern | rvba, hopefully this should give me a local maas that can work totally offline | 12:21 |
rvba | dimitern: yep, that's the goal of this tutorial. | 12:21 |
dimitern | rvba, cheers | 12:21 |
rvba | dimitern: please take notes about things that aren't clear enough… and we will improve the tutorial. I'm sure you're not the only one interested. | 12:22 |
dimitern | rvba, will do | 12:22 |
rvba | Ta | 12:22 |
dimitern | rvba, downloading from maas.ubuntu.com is very slow compared to archive.ubuntu.com btw - I managed to mirror 176 GB in about 4h, but for ~14GB from maas.ubuntu.com it already takes more than 20m | 12:23 |
maasnoob | Hello all | 12:26 |
maasnoob | I've got an interesting question on MAAS | 12:26 |
maasnoob | I've been looking at playing around with MAAS for building cloud networks but is there any reason why this couldn't been used to build a local network as well? | 12:26 |
mpontillo | hey dimitern, I'm awake now ;-) | 13:36 |
dimitern | mpontillo, hey :) | 13:39 |
dimitern | mpontillo, I've followed your local test/dev completely offline maas setup | 13:39 |
dimitern | mpontillo, and have a few question if you have time | 13:39 |
mpontillo | dimitern: I'm just using Apache to serve up the files for the mirror | 13:44 |
dimitern | mpontillo, yeah, I've configured nginx for the same purpose | 13:45 |
mpontillo | dimitern: cool. I have a symlink images -> /home/mirror/maas/images and I do most of my mirroring as the 'mirror' user | 13:45 |
mpontillo | from /var/www/html in my case | 13:45 |
dimitern | mpontillo, I'm having a *really* slow download from maas.ubuntu.com for sstreams-mirror - is this the same for you as well? | 13:45 |
dimitern | mpontillo, in contrast archive.u.c mirror completed in 4h (all 180GB of it) | 13:46 |
mpontillo | let me check dimitern | 13:46 |
dimitern | mpontillo, also I suspect not all of the ephemeral-v2 images are needed (shouldn't only the trusty ones work just as well?) | 13:46 |
mpontillo | dimitern: I'm running: sstream-mirror -v --max 1 http://maas.ubuntu.com/images/ephemeral-v2/releases/ $HOME/maas-test/images/ephemeral-v2/releases | 13:52 |
mpontillo | dimitern: have grabbed 500 MB so far | 13:52 |
mpontillo | my complete mirror of daily/ and releases/ is 17GB, FYI | 13:53 |
mpontillo | dimitern: here's the script I'm using for MAAS images mirroring now: http://paste.ubuntu.com/11793197/ | 13:54 |
dimitern | mpontillo, my sstreams-mirror just finished (after about 3h) fetching 15GB | 13:54 |
mpontillo | do you have the --max 1 in there? in my original talk, I forgot to put that in there. | 13:58 |
mpontillo | since it's so slow, I guess you can see why I'd rather have a local mirror than hit the server each time ;-) | 13:59 |
mpontillo | I'll let you know how long mine took when it completes. | 13:59 |
dimitern | mpontillo, I didn't, but now I added it | 13:59 |
mpontillo | dimitern: if you run it again, it should trim down some of that disk usage.. and you can also grab daily/ to test with =) | 14:00 |
dimitern | mpontillo, it's running now, will see how long it takes :) | 14:02 |
dimitern | mpontillo, oh, sstream-mirror is not so dumb as I thought - it detected I've already had all releases and didn't try downloading them again; dailies are now being downloaded | 14:05 |
mpontillo | dimitern: yeah, once you have the mirror going, it's pretty painless to maintain. I just have a cron job run the script a few times per day to check for new images | 14:07 |
mpontillo | I've grabbed ~4 GB so far | 14:08 |
dimitern | mpontillo, I'll set it up similarly | 14:12 |
dimitern | mpontillo, what's the --max 1 argument for? | 14:12 |
mpontillo | dimitern: I think you get historic images without it.. --max 1 just gives you the latest | 14:13 |
dimitern | mpontillo, I see, ok | 14:14 |
gnuoy | I'd like to add power support for a new power type. Are there any guides or pointers out there? If not I'll dig around in the code (maas 1.8). | 14:24 |
rvba | gnuoy: I can find a commit that does just that… hang on… | 14:25 |
gnuoy | that would be fantastic, thank you | 14:25 |
rvba | gnuoy: bzr log -p -r 3663 | less | 14:26 |
mpontillo | gnuoy: rvba may have beat me to it, but here's the complete *removal* of a power type (and corresponding probe-and-enlist code) https://code.launchpad.net/~mpontillo/maas/bug-1451852-remove-legacy-esxi-probe-and-enlist, which may be instructive | 14:28 |
gnuoy | rvba, mpontillo, perfect, thanks | 14:29 |
mpontillo | gnuoy: rather, I meant to link to this, which will give you the diff https://code.launchpad.net/~mpontillo/maas/bug-1451852-remove-legacy-esxi-probe-and-enlist/+merge/259075 | 14:29 |
mpontillo | dimitern: just completed; 9.0 GB in 48 minutes | 14:40 |
dimitern | mpontillo, yeah, I guess it's faster in the US | 14:40 |
dimitern | mpontillo, mine is still going, but I'll leave it running | 14:40 |
mup | Bug #1469742 opened: Failure to start sm15k servers because of exception to set power-type: KeyError: u'sm15k' <oil> <MAAS:Incomplete> <https://launchpad.net/bugs/1469742> | 14:40 |
kiko | caribou, hmm, so you want to tell the system to somehow boot from the second disk? | 14:41 |
* mpontillo does a traceroute; looks like the servers are in the UK (~160 ms away from me) | 14:41 | |
kiko | I don't think we support that | 14:41 |
caribou | kiko: well, that's my question : does Maas keeps track of the booting disk ? | 14:42 |
caribou | kiko: I thought of zapping the grub MBR off the first disk but no luck so far | 14:43 |
=== jfarschman is now known as MilesDenver | ||
kiko | caribou, I do not think it does | 15:20 |
kiko | caribou, first, I don't know if we even know what disk is the system disk | 15:21 |
kiko | but second, I don't know if we have a way to tell the node to boot off a specific disk | 15:21 |
caribou | kiko: that's what I thought; my guess is that it's the first grub MBR found that kicks the boot | 15:21 |
caribou | kiko: what I need to find is how to get the second disk to be the only one with a valid grub MBR to boot from | 15:22 |
caribou | well, if there is only one disk, my curtin partitioning works | 15:22 |
caribou | so that's what I'm blogging about | 15:22 |
kiko | I don't know how that phase is booting is handled | 15:23 |
kiko | and I'm not sure it is documented either | 15:23 |
kiko | rvba, roaksoax: do either of you know? | 15:23 |
mup | Bug #1441404 changed: 1.8b1 Accessing nodes detail page gets stuck in Loading <oil> <MAAS:Incomplete> <https://launchpad.net/bugs/1441404> | 15:59 |
caribou | kiko: well, I'm documenting a single disk partitioning for now. I'll add another post if I figure out multi-disk | 16:16 |
mup | Bug #1441404 opened: 1.8b1 Accessing nodes detail page gets stuck in Loading <oil> <MAAS:Incomplete> <https://launchpad.net/bugs/1441404> | 16:20 |
mup | Bug #1441404 changed: 1.8b1 Accessing nodes detail page gets stuck in Loading <oil> <MAAS:Incomplete> <https://launchpad.net/bugs/1441404> | 16:23 |
mup | Bug #1469822 opened: Updating a node fails with invalid OS and distro series <landscape> <MAAS:New> <https://launchpad.net/bugs/1469822> | 17:53 |
mup | Bug #1469846 opened: UCS chassis enlist Failed to probe and enlist UCS nodes: list index out of range <MAAS:New> <https://launchpad.net/bugs/1469846> | 18:56 |
busterswt | Has anyone had success accessing the complete MAAS GUI behind a SOCKS proxy? Some pages don't load completely, and it seems to hang on the yui URIs | 19:09 |
bdx | whats going on everyone? I just upgraded to maas 1.8 under trusty....I am experiencing an issue where I cannot save edits to a nodes details e.g. hostname or arch or any editable field for a node | 20:26 |
bdx | I'll fire up a bug report here asap....just thought I would check in and see if there might be a reason for this what seems like read only gui config | 20:27 |
mup | Bug #1469874 opened: MAAS 1.8 not allowed to edit node details in gui after upgrade <maas> <maas-gui> <MAAS:New> <https://launchpad.net/bugs/1469874> | 20:50 |
=== jfarschman is now known as MilesDenver |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!