[09:32] Bug #1526532 opened: [xenial, 1.10] enlistment failing [11:00] Any update with the 1.9.2 package? [11:31] I have a node that doesn't seem to get past Commissioning, using MAAS 2.0.0. The node is PXE booting and getting an IP. I have verified that the node is reachable via ping, and even attempted to ssh into it (public key denial). Are there any logs I can check on the node about it's progress? [13:03] Bug #1562249 opened: Failed to deploy machine with HP Smart Array Raid 6i [13:45] Bug #1526532 changed: [xenial, 1.10] enlistment failing [13:45] Bug #1562249 changed: Failed to deploy machine with HP Smart Array Raid 6i [13:57] Bug #1562249 opened: Failed to deploy machine with HP Smart Array Raid 6i [14:18] Bug #1574113 changed: No way to inject apt archive/mirror key 'in-target' before apt starts processing packages [17:03] mpontillo, roaksoax, blake_r: hey guy, are you aware of the issues with deploying nodes commissioned with biosdevname-enabled releases (e.g. xenial) and having too long interface names (e.g. 'enxxaabbccddeef0' for a usb2eth second NIC)? [17:03] s/guy/guys/ [17:05] maas (and curtin) render /e/n/i in this case which cannot be applied, as creating a VLAN on such an interface (e.g. 'enxxaabbccddeef0.1234') exceeds the device name limit of 16 characters [17:05] I'm filing a bug now for that [17:22] https://bugs.launchpad.net/maas/+bug/1574771 [17:22] Bug #1574771 opened: MAAS/curtin generate invalid /e/n/i and failed deployment for nodes with long (biosdevname) interface names, which in turn have VLANs [17:26] dimitern: thanks, good find. Hadn't seen that before; will look. [18:19] Hi any can help whit my build? [18:19] is very simple [18:20] i can build ubuntu openstack private cloud whit 2 phisican servers? [18:20] phisical* [18:25] Bug #1574771 changed: MAAS/curtin generate invalid /e/n/i and failed deployment for nodes with long (biosdevname) interface names, which in turn have VLANs [19:43] Bug #1571898 changed: [2.0 beta 2] Fabric's are not created sequencially [21:09] quick question - I've created a custom image in maas, and it shows up under the generated images section - but it doesn't appear in the dropdown for deployment [21:09] is there an additional step I need to do to ready an image for deployment? [21:11] LiftedKilt: has it been synced to the cluster controller ? [21:11] roaksoax: aha no - the cluster-master is saying it is out of sync [21:11] they are on the same machine - how do I force a sync? [21:12] LiftedKilt: is there an error of why that might be ? no issues in the logs ? [21:14] exceptions.IOError: Unable to open http://localhost:5240/MAAS/images-stream/custom/amd64/generic/rancheros0.4.4/20160425/root-tgz. mirrors=[] [21:15] LiftedKilt: so that's the error...now i wonder what the underlying issue might be [21:15] LiftedKilt: do you have enough disk space? [21:16] roaksoax: yeah I've got plenty of space [21:16] roaksoax: I'm going to try restarting maas-* [21:19] nope. didn't fix it [21:20] LiftedKilt: the weird thing is that it is an IOError, so i wonder if it is related to the image itself or something else [21:20] it might be - it's an image for rancherOS [21:21] LiftedKilt: never tested one of those [21:21] images should be in tar.gz? [21:22] it's available in tar.gz and iso [21:22] LiftedKilt: that's probably the reason why then [21:22] LiftedKilt: it is probably [21:22] LiftedKilt: that [21:22] LiftedKilt: the images need to be in tgz's yes [21:23] the image i tried uploading was a tgz [21:23] I was just making sure that was correct [21:23] LiftedKilt: if you could file a bug, we'll be able to look at it [21:24] roaksoax: ok - I'll poke at it first for a bit. I might try on 2.0 and see how that works [21:25] LiftedKilt: to be fair that codepath is fairly unchanged [21:25] roaksoax: ok - thanks