| Odd_Bloke | bjf: Heads up: argyle is running in PS4 at the moment, and we've been asked to move it to PS4.5 by the end of the week. | 08:13 |
|---|---|---|
| Laney | did aufs go away in our 4.0? | 09:29 |
| Laney | my schroots broke :( | 09:29 |
| Odd_Bloke | bjf: apw: Oh, it's not by the end of the week, it's by July 15 (next Wednesday). | 09:34 |
| Odd_Bloke | bjf: apw: I've redeployed the service, and opened RT #82789 to open up the same firewall holes as we have for the PS4 deployment. | 09:35 |
| infinity | Laney: Erk, it sure did... | 09:35 |
| infinity | apw: Where'd aufs go? | 09:36 |
| infinity | Laney: You can s/aufs/overlay/ for now (or downgrade). | 09:36 |
| Laney | infinity: Ya, done that, thanks. | 09:38 |
| apw | infinity, it was meant to be there, i am sure tim updated it | 10:32 |
| infinity | apw: I'm sure modinfo tells me I don't have it. | 12:29 |
| infinity | Althought, CONFIG_AUFS_FS=m ... | 12:29 |
| infinity | Wha? | 12:29 |
| apw | infinity, yeah its disabled in the makefile, seems to build just fine, so i am proposing to do a non-abi bumper with it enabled | 12:52 |
| infinity | apw: If it works, that seems reasonable. | 12:53 |
| bjf | Odd_Bloke, so our "cloud provider" has asked us to move to a new cloud but it's up to us to do that? and i suppose any/all firewall exceptions are also up to us to tell them about? | 13:45 |
| rbasak | awe: do you mind posting your /var/log/apt/history.log somewhere from yesterday's grub2 incident? I'd like to see what exactly did what when your system originally broke to understand the issue better. | 13:49 |
| awe | rbasak, in a mtg, and just about to join a standup; I'll ping you when I free up | 13:54 |
| rbasak | OK, thanks. | 13:54 |
| Odd_Bloke | bjf: Something like that, yes. | 14:02 |
| Odd_Bloke | bjf: But they have at least been fairly responsive; new rabbitmq is at 162.213.33.247 and the firewall holes should have been punched. | 14:03 |
| Odd_Bloke | (In their defense, all I had to do was say "copy the PS4 firewall rules", so hopefully they should all be correct) | 14:03 |
| bjf | Odd_Bloke, since we access it via ip addr, it will require code changes | 14:05 |
| Odd_Bloke | bjf: Yeah. :( | 14:06 |
| bjf | Odd_Bloke, from what you are telling me. the new one is in place and i should be able to switch over right now if i wanted ... correct? | 14:07 |
| Odd_Bloke | bjf: That's my understanding; but check first? | 14:08 |
| bjf | Odd_Bloke, ok ... i'll do a little checking and if it goes well, make the switch | 14:09 |
| bjf | apw, ^^^^ i'll do some testing to make sure the msgq service is working for the systems we want it | 14:13 |
| apw | bjf, sounds good, as i will need to shock-and-awe on the dashboard | 14:16 |
| awe | did someone say my name? ;)- | 14:18 |
| apw | :) | 14:18 |
| apw | infinity, ok just enablnig aufs seems to work just fine, soooo ... i'll get that uploaded | 14:18 |
| apw | ogasawara, seems we are missing aufs in the kernel we just did, so i am proposing to do a non-abi bumper for it | 14:19 |
| awe | rbasak, http://pastebin.ubuntu.com/11848708/ | 14:19 |
| ogasawara | apw: ack | 14:19 |
| infinity | apw: Works for me. | 14:19 |
| infinity | apw: I can never decide from minute to minute if my schroots should be overlay, aufs, or lvm, but hey, it's nice to be spoiled for choice. | 14:20 |
| infinity | apw: (I still wish overlay could paper over the one or two things that make me keep turning to aufs...) | 14:20 |
| rtg | apw, the 4.0 Wily upload is missing AUFS ? | 14:21 |
| infinity | rtg: Yep. | 14:21 |
| infinity | rtg: It's in the config, but apparently disabled in the Makefile. | 14:21 |
| infinity | (oops) | 14:21 |
| rtg | infinity, oh, damn | 14:21 |
| rtg | sure is | 14:22 |
| rtg | better check the 4.1 branch | 14:22 |
| rtg | looks like that one is OK | 14:23 |
| rbasak | awe: thanks! | 14:23 |
| infinity | rtg: Misplaced your cowboy hat in between versions? | 14:23 |
| awe | np | 14:23 |
| infinity | rbasak: So, that totally has apt-get being the culprit, not update-manager. We can stop panicking. | 14:24 |
| infinity | rbasak: As in, *someone* got impatient that grub wasn't upgrading and ran "apt-get install grub" and said "sure, remove everything" when it prompted. | 14:25 |
| infinity | In fact, wait. That's grub1, even. | 14:25 |
| infinity | awe: So, that was totally self-inflicted. | 14:25 |
| infinity | awe: Your manual installation of grub1 removed grub2, and you said "yes". | 14:26 |
| rbasak | infinity: AIUI, awe had a problem after the update, and he tried to fix it by installing grub | 14:26 |
| rbasak | infinity: but yes, the problem wasn't cause because the grub2 signed package got removed. Must have been caused by some other thing. | 14:26 |
| infinity | rbasak: Well, there are no removals until that point. | 14:26 |
| rbasak | Right | 14:26 |
| awe | infinity, I only did the manual install of grub *after* my system stopped rebooting correctly | 14:27 |
| awe | and yes, it was a mistake | 14:27 |
| awe | but didn't really worsen or improve anything | 14:28 |
| infinity | awe: Check. What are/were the symptoms of "not booting correctly"? | 14:28 |
| apw | rtg, yes, you ripped aufs out and switch the way it was updated for v4.1 | 14:29 |
| awe | I'm dual-booting a macair osx + utopic | 14:29 |
| awe | after yesterday morning's update | 14:29 |
| rtg | apw, right, I'm using the upstream mechanism | 14:29 |
| awe | my machine no longer booted into rEFInd | 14:29 |
| awe | I got a "Your screwed screen" | 14:29 |
| awe | ( ie. do you wanna re-install OS X from TimeMachine backup, fresh install, go online for help, or bring up disk tools ) | 14:30 |
| apw | rtg, indeed, i mean that is why the disablement didn't get carried over, one was against ubuntu/Makefile, the other against fs/Makefile | 14:30 |
| rtg | apw, yup | 14:30 |
| awe | so now on restart, I need to do an Option+Pwr boot, and then select EFI boot | 14:30 |
| awe | the other options being the two OSX recovery partions | 14:30 |
| infinity | awe: Curious. I don't know much of anything about rEFInd, but since rbasak's been trying to be helpful, I'll let him keep at it. :P | 14:31 |
| * awe heads to yet another mtg | 14:31 | |
| awe | thanks infinity | 14:31 |
| rbasak | Well, I'm not sure any more :-/ | 14:31 |
| rbasak | Not sure how much more helpful I can be. I was concerned yesterday that there might be a grub2-signed breakage problem, and so when awe's problem appeared, was concerned that it was that. | 14:32 |
| rbasak | But it wasn't, so perhaps my concern isn't as big of an issue (that's the trouble with a sample size of 1). But that doesn't help at all with awe's actual problem, of which I have no idea. | 14:32 |
| arges | rodsmith: hey, awe had some issues with rEFInd. Maybe you can help? | 14:36 |
| rodsmith | Hi. I'm told somebody has a rEFInd question. I'm the current maintainer. What's the question? | 14:36 |
| rtg | apw, there are significant changes to overlayfs in 4.2; perhaps you could add to your TODO list to forward port "UBUNTU: SAUCE: overlay: add backwards compatible overlayfs format support V3" | 14:36 |
| apw | rtg, ack | 14:36 |
| rtg | apw, in unstable master | 14:36 |
| apw | master or master-next ? | 14:36 |
| awe | rodsmith, rbasak, I just joined a planning meeting; short version... I updated yesterday, and my macair no longer boots into rEFInd | 14:37 |
| rtg | apw, I think I'm gonna drop unstable master-next 'cause it really isn't very useful. what do you think ? | 14:37 |
| awe | now, I have to do an 'Option' + Pwr, and select "EFI Boot" every time | 14:37 |
| awe | not sure if I re-bless from OS X would fix this? | 14:37 |
| awe | This is the first time an Ubuntu update has broken things for me | 14:37 |
| rodsmith | awe: Can you get into both OS X and Ubuntu via the Option key? | 14:37 |
| awe | I get to rEFInd via the option key, and have only booted Ubuntu from there since | 14:38 |
| awe | haven't checked OSX | 14:38 |
| rodsmith | awe: If you can get into OS X, try re-installing rEFInd. | 14:38 |
| awe | without the option key | 14:38 |
| awe | I get a "Do you want to re-install or fix things" screen | 14:38 |
| rodsmith | awe: The Ubuntu update probably tried to install/re-install GRUB, which messed things up. | 14:38 |
| awe | right | 14:39 |
| awe | I'll try a re-install from OS X and see if that fixes things | 14:39 |
| rodsmith | awe: The "re-install or fix things" prompt sounds like APT didn't quite set everything up right. | 14:39 |
| awe | right | 14:39 |
| awe | thanks | 14:39 |
| rtg | manjo, smb: I see that dm-raid45 hasn't been updated since April 2009. We are carrying it as a SAUCE driver. Do you think it is still relevant ? | 15:19 |
| manjo | rtg, wow... I will need to look coz I am not sure who is using that | 15:20 |
| smb | rtg, depedns on how many people use a non-Intel fakeraid in bios mode... | 15:20 |
| smb | Intel MSM and DDS should be handled by mdadm nowadays | 15:21 |
| smb | rtg, Sure we will notice (after release) when you rip it out for now | 15:21 |
| rtg | smb, well, I haven't ripped it out yet. I'm just reviewing SAUCE patches | 15:22 |
| manjo | smb, wasint there an issue where softraid was not write able ? | 15:22 |
| manjo | smb, May be that issue with mdadm is already fixed | 15:22 |
| smb | manjo, at some point long time ago. but afaict Trusty works without dmraid | 15:23 |
| manjo | rtg, will that impact backport kernels for lts ? | 15:23 |
| rtg | manjo, yup | 15:24 |
| manjo | ugh | 15:24 |
| rtg | maybe it should wait until after 16.04 to be removed | 15:24 |
| manjo | when does precise sunset? /me does the math | 15:24 |
| manjo | 12.04 .. so 17.04? | 15:25 |
| rtg | yup | 15:25 |
| manjo | s**t | 15:25 |
| smb | rtg, Might be that is what we said before T... But with a crisp memory like mine and we should have written it down... but then maybe we did and I just forgot where | 15:26 |
| rtg | smb, its no big deal. so far maintenance has been easy, but I wanted to be sure the driver actually still worked | 15:27 |
| TJ- | rtg: If it's any help I can do some tests with a Promise controller; which kernel/ubuntu release would you want testing? | 15:29 |
| smb | rtg, I doubt I ever tested the other module ever again after mdadm worked for Intel... and none of my other hw would have fakeraid beyond levels 1 | 15:29 |
| manjo | rtg, also people using vmware stuff might still need that module on precise | 15:29 |
| rtg | ok, I'll defer removing the driver | 15:29 |
| apw | rtg, we if we are removing it in 16.04 we should disable it in the main build and re-enable it in the lts-backport | 15:47 |
| rtg | apw, I was thinking _after_ 16.04 | 15:47 |
| apw | then you might need to support it in the backports up to and including 18.04 in 16.04 even if it is disabled in those releases | 15:53 |
| apw | because it _is_ enabled in 16.04 | 15:53 |
| rtg | apw, 16.04 is the last backport for 14.04. perhaps we should only enable dm-raid45 in the backport and not in 16.04 proper | 15:55 |
| apw | rtg, that was my suggestion indeed | 15:57 |
| infinity | rtg: I concur with apw (and you). If you intend to drop drivers, do so in 16.04, but add them back to the lts-x kernel. | 16:09 |
| rtg | infinity, now I just gotta remember to do it | 16:09 |
| infinity | rtg: Though, you could also drop them in wily and add them back to the lts-w kernel too. | 16:09 |
| infinity | rtg: Which gives people some time to get used to the idea. | 16:09 |
| rtg | yup | 16:10 |
| infinity | rtg: Makes an argument for maintaining lts-w starting around, oh, now, though. So you don't end up with 73 post-its detailing "stuff that needs to happen on top of the backport". :) | 16:11 |
| rtg | thats just too logical | 16:11 |
| unixabg | Greetings, does the 15.04 kernel have multi-layer overlayfs support? (multi-layer meaning more than 2) | 16:12 |
| apw | unixabg, that is a feature of later kernels, so wily may have it iwth the very latest kernels | 16:15 |
| unixabg | apw: humble appreciation for response. | 16:19 |
| apw | unixabg, i don't think 3.19 had it, but i am not 100% sure | 16:20 |
| unixabg | apw: would you know what version would have such support? | 16:21 |
| unixabg | I ask since I am wanting to test it some and I need partial .squashfs files for service packs. | 16:22 |
| apw | unixabg, looks like v4.0 was the first upstream release with it in | 16:23 |
| apw | so only wily very latest kernel would have it | 16:23 |
| unixabg | apw: thank you and I will work on testing with that then. | 16:28 |
| unixabg | apw: what is the shortest route to install and test the 4.x kernel? | 16:31 |
| apw | well other than running wily, you could install that kernel from the archive by hand | 16:31 |
| unixabg | apw: thank you for the assistance I will download from http://cdimage.ubuntu.com/daily-live/current/ | 17:12 |
| dsd | hi, i read that Ubuntu Wily might be based on Linux 4.1. is there a Ubuntu kernel git tree for that already? | 18:59 |
| rtg | dsd, ultimately Wily will release with 4.2 | 19:03 |
| dsd | i see | 19:06 |
| dsd | and have you started the process of moving the patches from 3.19 to a newer kernel like that? | 19:07 |
| dsd | or is that going to happen a bit later | 19:07 |
| rtg | dsd, All 4.1 work is happening here until 4.2 is released: git://git.launchpad.net/~ubuntu-kernel/ubuntu/+source/linux/+git/wily master-next | 19:08 |
| dsd | thanks! | 19:11 |
| unixabg | apw: have you worked on casper? I ask since I see an apw there on a commit message. | 20:01 |
| apw | unixabg, only in a very peripheral way, but the last upload was in my name in trusty yes | 21:13 |
| SturmFlut | Good midnight! Anybody still awake? | 22:24 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!