thai_ | anyone got experiences with openstack on 16.04 multi nodes? | 00:11 |
---|---|---|
=== sikun_ is now known as sikun | ||
RoyK | thai_: We have a 16-node openstack cluster (that is 16 compute nodes, routers and storage are separte) at work. I think it's till on 14.04. I could ask around if you have a specific question | 01:15 |
Zeranoe | I'm setting up a dev lamp server. How should I go about granting permissions for my user and the web server to write to /var/www/html? | 03:33 |
Zeranoe | I'm thinking maybe I should make www-data the owner of the group and add my user to that group? | 03:39 |
sarnold | Zeranoe: that's one decent way; I dislike the idea that the server may have the ability to write to those files, so I'd also suggest that www-data itself not own those files; either group or world permissions to read them is better, imho | 05:04 |
=== lifeless_ is now known as lifeless | ||
Zeranoe | sarnold: So maybe don't touch own, but add the user to www-data and grant write permissions for groups for /var/www/html ? | 05:37 |
Zeranoe | I guess I'm not seeing how that's any different... | 05:38 |
sarnold | Zeranoe: I just like to make sure that the web server itself can only write to its log files and database / fpm sockets if you can | 05:40 |
Zeranoe | sarnold: Doesn't that cripple a large number of web applications...? | 05:40 |
sarnold | Zeranoe: if your apps have to allow uploading stuff, maybe that too :) | 05:41 |
Zeranoe | I think a better solution would be to have decently written code so that rouge writes don't happen | 05:42 |
sarnold | somehow 'write bug-free code' just never gets to the top of manager's priorities :( | 05:47 |
Zeranoe | I hear that | 05:47 |
xmj | "write bug-free code" never gets on employee priorities either | 06:57 |
xmj | or rather, results | 06:57 |
xmj | so why bother | 06:57 |
Odd_Bloke | sarnold: hallyn: Thanks. :) | 08:46 |
ajikny | tomreyn: there? | 08:56 |
ajikny | I have got the kernlog and the syslog | 08:56 |
ajikny | from the server | 08:56 |
ajikny | I have a weird problem. I upgraded ubuntu server 12.04 LTS to 1.404 LTS. | 08:58 |
ajikny | 14.04 LTS | 08:58 |
ajikny | as soon as i upgraded the keyboard and network stopped working | 08:59 |
ajikny | The keyboard works till we load grub. | 08:59 |
ajikny | as soon as the kernel is load the kb stops working | 08:59 |
ajikny | funnily enough the kb and network both work in the live boot | 09:00 |
ajikny | i have tried installing image-extra-generic, update-usbids, linux-firmware, linux-firmware-nonfree followed by update-initramfs -k all -c | 09:01 |
zioproto | Hello. I have a weird problem I am building deb packages for ubuntu trusty, the control file in the deb package is broken. In the depends section some other packages have an extra dash - at the end, and this makes the package impossible to install | 09:07 |
zioproto | https://code.launchpad.net/~zioproto/ubuntu/+source/cinder/+git/cinder/+merge/304341 | 09:07 |
zioproto | anyone had a similar problem before ? | 09:10 |
Odd_Bloke | zioproto: Can you paste an example of the problematic file? | 09:18 |
zioproto | Odd_Bloke: check the comment here https://code.launchpad.net/~zioproto/ubuntu/+source/cinder/+git/cinder/+merge/304341 | 09:19 |
zioproto | I pasted the file | 09:19 |
Odd_Bloke | zioproto: Oh, apologies. | 09:20 |
zioproto | no problem | 09:20 |
zioproto | the package where I have the problem is python-cinder | 09:20 |
zioproto | di you see that python-novaclient- ? the ending - in the control file brakes the dependencies | 09:21 |
Odd_Bloke | zioproto: Yeah, that's strange. | 09:22 |
Odd_Bloke | zioproto: How are you building your .dsc? (And can you paste it?) | 09:23 |
zioproto | sure, give me a second | 09:23 |
zioproto | here is the dsc http://pastebin.com/2jCg74gu | 09:24 |
Odd_Bloke | zioproto: Hmm, not really sure what the problem is I'm afraid. | 09:25 |
ajikny_ | I have a weird problem. I upgraded ubuntu server 12.04 LTS to 14.04 LTS as soon as i upgraded the keyboard and network stopped working The keyboard works till we load grub. as soon as the kernel is load the kb stops working [14:30] funnily enough the kb and network both work in the live boot i have tried installing image-extra-generic, update-usbid | 09:27 |
zioproto | sorry I have to be away from keyboard for 30 mins, I will be back | 09:32 |
ikonia | do you see anything in /dev/input | 09:32 |
ikonia | do you see anything in the syslog about the network devices being loaded/brought online or failing | 09:32 |
ajikny_ | ikonia: i get by-id by-path event [0-4] js0 mice mouse0 mouse1 | 09:33 |
ikonia | where do you get that ? | 09:33 |
ajikny_ | i have livebooted inot the system | 09:33 |
ajikny_ | *into | 09:34 |
ikonia | thats not really going to help diagnose the problem | 09:34 |
ajikny_ | i have the chrooted dmesg and sys logs | 09:34 |
ikonia | dmesg isn't a file | 09:35 |
ajikny_ | if it can help | 09:35 |
ikonia | it's real time information | 09:35 |
ikonia | so thats not going to help getting dmesg from the livecd | 09:35 |
ikonia | I assume you have no boot splashes enabled when the system boots | 09:35 |
ajikny_ | no | 09:35 |
ikonia | when the system boots do you see any errors/warnings, especailly around the network | 09:35 |
ajikny_ | it stops at waiting for network configuration | 09:36 |
ajikny_ | for 60seconds and then proceeds with boot without full config | 09:36 |
ikonia | are you using dhcp ? | 09:36 |
ajikny_ | no.. static config | 09:36 |
ikonia | ok - so the syslog from the system should show it trying to bring up the devices | 09:37 |
ikonia | find that in the syslog and see what it's doing with the network | 09:37 |
ajikny_ | okayyy... hold on | 09:37 |
zioproto | Odd_Bloke: I am trying to rebuild the package without my patch to see if I have anyway the problem with the control file | 09:43 |
zioproto | Odd_Bloke: James Page from canonical commented on the merge request saying I am maybe hitting a known bug of cloud-archive-utils | 10:06 |
zioproto | and yes I am able to reproduce the problem even with the upstream package without my patch | 10:06 |
zioproto | jamespage: oh you are here :) | 10:06 |
jamespage | zioproto, hey! | 10:06 |
jamespage | zioproto, just digging on your problem now | 10:07 |
zioproto | jamespage: to push the fixed changelog, should I just commit amend and push force, or you prefer a tree with multiple commits ? | 10:07 |
zioproto | jamespage: thank you ! | 10:07 |
jamespage | zioproto, amend and force push is good for me | 10:07 |
jamespage | zioproto, can you confirm your cloud-archive-utils pkg version? | 10:08 |
zioproto | jamespage: Installed: 0.1-0~50~ubuntu16.04.1 | 10:09 |
jamespage | zioproto, ok that's good | 10:09 |
* jamespage thinks | 10:09 | |
zioproto | jamespage: check the merge request page, I added more stuff | 10:09 |
zioproto | jamespage: my doubt is, when I use the command sbuild-liberty, then in the chroot what version of could-archive-utils will be used ? | 10:10 |
jamespage | zioproto, for reference you can drop "-d trusty-amd64" from your build line - sbuild-liberty deals with that | 10:10 |
jamespage | zioproto, it should pick a trusty schroot | 10:10 |
jamespage | zioproto, schroot -l might be helpful here | 10:10 |
zioproto | I have a two lines output | 10:11 |
jamespage | zioproto, there is another possibility - do you keep your schroots up-to-date? | 10:11 |
zioproto | chroot:trusty-amd64 | 10:11 |
zioproto | source:trusty-amd64y | 10:11 |
jamespage | zioproto, that's fine - no need for -d at all then | 10:11 |
zioproto | how can I update the chroot ? | 10:11 |
zioproto | I think this all vm where I am working is at most two weeks old | 10:12 |
jamespage | zioproto, well it should be ok | 10:13 |
jamespage | lemme dig out my update script | 10:13 |
jamespage | zioproto, http://paste.ubuntu.com/23111240/ | 10:14 |
jamespage | I have that on cron.daily | 10:14 |
jamespage | zioproto, could you pastebin the entire build log? | 10:15 |
zioproto | jamespage: 0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded. | 10:15 |
zioproto | looks like it was upgraded already | 10:15 |
zioproto | jamespage: yes, do I have to screen scrape it or it is automatically saved in a file ? | 10:16 |
jamespage | zioproto, its saved to a file locally | 10:16 |
jamespage | zioproto, we had a bug in dh-python todo with dependency parsing which this looks like | 10:16 |
jamespage | but I'm not 100% convinced | 10:16 |
jamespage | https://launchpad.net/ubuntu/+source/dh-python/1.20140128-1ubuntu8.2 | 10:16 |
jamespage | as I fixed that | 10:16 |
zioproto | I guess you want this file right ? | 10:17 |
zioproto | /home/ubuntu/ubuntu/logs/cinder_7.0.2-0ubuntu1_amd64-20160830-0942.build | 10:17 |
zioproto | it is a symlink in the folder where I did the build | 10:17 |
zioproto | jamespage: can you download it from here ? https://www.dropbox.com/s/y8u3m1sxog39ovo/cinder_7.0.2-0ubuntu1_amd64-20160830-0942.build?dl=0 | 10:19 |
jamespage | zioproto, hmm | 10:20 |
jamespage | dh-python_1.20140128-1ubuntu8 | 10:20 |
jamespage | that looks wrong | 10:20 |
zioproto | how can I update that ? | 10:21 |
jamespage | zioproto, well the sbuild-liberty wrapper should do that | 10:22 |
ajikny_ | ikonia: read the logs, the network manager tries to do a dhcp request | 10:23 |
ajikny_ | ikonia: this is my first time reading logs and understanding so sorry it took some time to understand. I am willing to learn and get my hands dirty tough | 10:23 |
zioproto | jamespage: so I did not do anything wrong :) | 10:24 |
jamespage | zioproto, no - just trying to figure out why you schroot is not picking that update up - mine is! | 10:24 |
jamespage | so is our backport-o-matic for the UCA | 10:25 |
jamespage | grrr | 10:25 |
zioproto | what is backport-o-matic ? | 10:25 |
=== Geom` is now known as Geom | ||
ajikny_ | Respected all | 10:27 |
ajikny_ | I got a really weird problem on upgrading ubuntu 12.04 to 14.04 | 10:28 |
ajikny_ | after the upgrade the keyboard an network stopped working. | 10:28 |
ajikny_ | I cannot ssh into the system and i cannot type anything | 10:29 |
ajikny_ | The keyboard and network work in liveboot | 10:29 |
ajikny_ | Also keyboard works untill the grub menu. Once linux starts loading the wired usb keyboard stops working. The leds turn off | 10:30 |
jamespage | zioproto, its the automated process of backporting openstack related packages to the ubuntu cloud archive | 10:34 |
jamespage | it uses cloud-archive-utils as well | 10:34 |
jamespage | infact that's what cloud-archive-utils was written for | 10:35 |
zioproto | jamespage: how can I get a shell into the chroot using the schroot command to debug | 10:35 |
zioproto | ? | 10:35 |
zioproto | got it schroot -c trusty-amd64 | 10:38 |
zioproto | in the chroot I have this candidate for dh-python Candidate: 1.20140128-1ubuntu8 | 10:39 |
zioproto | jamespage: I think the problem is in the chroot I have only trusty/main and not trusty-updates/main | 10:40 |
jamespage | zioproto, yes that's the problem | 10:42 |
jamespage | zioproto, not quite sure how you managed to create a schroot like that tho | 10:42 |
jamespage | zioproto, how did you create the schroot? | 10:42 |
zioproto | following the ubuntu docs :) | 10:42 |
zioproto | hold on | 10:42 |
zioproto | I will show you | 10:42 |
zioproto | I use this script https://github.com/zioproto/ubuntu-cloud-archive-vagrant-vm/blob/xenial/bootstrap.sh | 10:43 |
zioproto | jamespage: look at this documentation page at step 7, there is SKIP_UPDATES=1 https://wiki.ubuntu.com/SimpleSbuild | 10:44 |
zioproto | jamespage: could that be the problem ? | 10:44 |
jamespage | zioproto, yeah - I'd drop | 10:44 |
jamespage | SKIP_UPDATES="1" | 10:44 |
jamespage | SKIP_PROPOSED="1" | 10:44 |
jamespage | SKIP_UPDATES stops you getting any updates for packages in 14.04 | 10:45 |
jamespage | which you need | 10:45 |
zioproto | OK, I will update my scripts | 10:45 |
jamespage | proposed is SRU's in test for 16.04 | 10:45 |
zioproto | I have to grab something to eat now, I will test asap and I will give you feedback. I will try to do it today. | 10:46 |
zioproto | are you going to be around in IRC for the rest of the day ? | 10:46 |
zioproto | thanks for the help. talk to you later | 10:47 |
jamespage | zioproto, yeah I'll be around for another 6 hrs or so | 11:01 |
LuisM | hi folks | 12:05 |
Ussat | o/ | 12:08 |
LuisM | Ussat: how to deal with memory recognize in Ubuntu Server inside ESXi? | 12:13 |
Ussat | not sure what ya mean | 12:13 |
LuisM | i want to do ubuntu hot add memory | 12:13 |
LuisM | https://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=1012764 | 12:13 |
LuisM | look | 12:13 |
LuisM | acpi_memhotplug | 12:14 |
Ussat | I have never hot-added memory in ESXi | 12:14 |
Ussat | in my experiance never works right | 12:14 |
patdk-wk | why wouldn't it work right? | 12:14 |
Ussat | X86 just isnt up to it imho | 12:14 |
patdk-wk | I have never had it not work | 12:14 |
patdk-wk | what does x86 have to do with it? | 12:14 |
patdk-wk | hot add/remove memory was designed for laptops | 12:14 |
patdk-wk | for power savings | 12:15 |
LuisM | yeah, i'm in x86, but my VMs are 64-bits | 12:15 |
Ussat | I always down my Linux systems to add mem or cpu or whatever, my AIX systems otoh, I DLPAR flawlessly | 12:16 |
ogra_ | patdk-wk, it surely works better oon all these s390x laptops than on x86 ... :P | 12:16 |
patdk-wk | ogra_, never had an issue doing it on linux, to add/remove ram | 12:16 |
Ussat | I know in theory it works on Power arch, but I have not tried it | 12:17 |
Ussat | maby will play with that a bit sometime | 12:17 |
Ussat | and by X86 I dont mean 32 or 64 bit, I mean X86 chipset | 12:18 |
LuisM | uhum | 12:24 |
LuisM | Ussat: in Solaris or AIX, hot memory add is flawlessly | 12:25 |
LuisM | :D | 12:25 |
Ussat | LuisM, I have not use Solaris in YEARS | 12:25 |
Ussat | I use AIX daily | 12:25 |
LuisM | aix > solaris | 12:26 |
LuisM | ;) | 12:26 |
Ussat | LuisM, but more and more I am doin all the Linux stuff here | 12:26 |
LuisM | same here | 12:26 |
patdk-wk | hot-add has always worked flawless in linux and ubuntu for me | 12:35 |
patdk-wk | dunno what the issue is | 12:35 |
patdk-wk | now, when you add memory, linux kernel doesn't enable that memory automatically, so you have to do that, or add a udev rule to do it for you | 12:35 |
patdk-wk | SUBSYSTEM=="cpu", ACTION=="add", TEST=="online", ATTR{online}=="0", ATTR{online}="1", RUN+="/usr/bin/logger onlining cpu: $env{DEVPATH}" | 12:36 |
patdk-wk | SUBSYSTEM=="memory", ACTION=="add", TEST=="state", ATTR{state}=="offline", ATTR{state}="online", RUN+="/usr/bin/logger onlining memory: $env{DEVPATH}" | 12:36 |
Ussat | patdk-lap, but in AIX or Solaris its automatic.....thats my point | 12:36 |
patdk-wk | and the kernel won't get the notification if you don't have the acpi_memhotplug module loaded | 12:36 |
patdk-wk | Ussat, heh? it is automatic here also | 12:36 |
Ussat | you dlpar memory,or a cpu or a pci slot... | 12:37 |
Ussat | <patdk-wk> now, when you add memory, linux kernel doesn't enable that memory automatically, so you have to do that, or add a udev rule to do it for you | 12:37 |
Ussat | you said it wasnt right there...... | 12:37 |
patdk-wk | maybe you mean, enabled | 12:37 |
patdk-wk | it's automatically added, it's not automatically enabled | 12:37 |
Ussat | I mean its automatically added and enabled | 12:37 |
patdk-wk | enabling it automatically, can be debated if that is a good idea | 12:37 |
Ussat | Not much to debate, you add it when you need it. | 12:38 |
Ussat | Or you have a system that automatically balances based on load | 12:38 |
patdk-wk | hmm, I add it to my laptop when I don't need it | 12:38 |
Ussat | laptop..... | 12:38 |
patdk-wk | I don't want to waste power in my laptop, except when I NEED the memory | 12:38 |
Ussat | ok I am talking real enterprise systems | 12:39 |
patdk-wk | this hot-add was designed and created for laptops | 12:39 |
Ussat | not little laptops | 12:39 |
patdk-wk | my laptop has 64gigs ram | 12:39 |
Ussat | I dont even bother with a *nix on a laptop or desktop | 12:39 |
Ussat | like I said, little laptop | 12:39 |
patdk-wk | your loss | 12:39 |
Ussat | Not at all | 12:39 |
Ussat | My enterprise systems have 512G RAM or better | 12:40 |
patdk-wk | still, I dunno of any laptops that can have >64gigs ram | 12:40 |
Ussat | and he was talking about ESXi, not a toy | 12:40 |
patdk-wk | no, he was talking about hot-add memory | 12:41 |
patdk-wk | that acpi interface was designed by and for laptops | 12:41 |
patdk-wk | the fact esxi uses it, doesn't matter | 12:41 |
patdk-wk | it's a power savings feature | 12:41 |
Ussat | No, it really isnt desinged as that, but believe what you eill | 12:42 |
Ussat | will | 12:42 |
patdk-wk | http://www.acpi.info/DOWNLOADS/ACPIspec-2-0a.pdf | 12:46 |
Ussat | I know what acpi is, and it was not desinged for hot-add memory......that was a later addition | 12:47 |
patdk-wk | heh? | 12:47 |
patdk-wk | how is a later addition, and being part of the specification matter? | 12:48 |
patdk-wk | it's part of the acpi standard | 12:48 |
patdk-wk | it doesn't matter if it was in predraft 0.1, or 1.0, or 2.0, or whatever, if it's in there it is a part of it | 12:48 |
Ussat | believe what you want, you are incorrect | 12:48 |
patdk-wk | well, go tell esxi to stop using acpi for hot-add memory | 12:49 |
Ussat | esxi uses an extension to acpi to accomplish that correctly, and it still has some major issues | 12:49 |
Ussat | and it can not be solved via software, its a limitation of the X86 chipset | 12:50 |
LuisM | patdk-wk: so, how to load that module in ubuntu-server if it doesn't exist? :( | 12:53 |
LuisM | what package do i need to install? | 12:53 |
LuisM | acpid? | 12:53 |
LuisM | or just add that udev rule? | 12:54 |
patdk-wk | it does exist | 12:54 |
patdk-wk | I just posted all the info above | 12:54 |
LuisM | yeah | 12:54 |
LuisM | but the module doesn't exist :( | 12:54 |
LuisM | i'm in 16.04 | 12:54 |
patdk-wk | hmm, haven't done 16.04 in a vm yet | 12:55 |
coreycb | ddellav, jamespage: I have changes coming for ceilometer in newton if you haven't already started those | 12:56 |
jamespage | coreycb, no - I was looking at keystone | 12:57 |
jamespage | coreycb, ddellav: did the syncs from exp this am btw | 12:57 |
patdk-wk | LuisM, there is no need | 12:57 |
coreycb | ddellav, jamespage: ah thanks. I'd like to wrap up the remaining non-client updates so that probably got some of them | 12:57 |
patdk-wk | it's built in on 16.04, it's not a module now | 12:57 |
patdk-wk | so only the udev rules would be needed to make it enabled automatically | 12:58 |
jamespage | coreycb, keystonemiddleware and auth1 need doing I think | 12:58 |
coreycb | jamespage, ok | 12:58 |
patdk-wk | CONFIG_ACPI_HOTPLUG_MEMORY=y | 12:58 |
LuisM | patdk-wk: hmmm, great | 12:58 |
ddellav | jamespage coreycb im working on keystonemiddleware atm | 12:58 |
LuisM | where did you see that? | 12:58 |
coreycb | ddellav, sweet thanks | 12:58 |
patdk-wk | in the kernel config file in /boot | 12:58 |
jamespage | coreycb, ddellav: seeing some crypto breaks in glance and barbican | 13:03 |
jamespage | something may be broken outside the direct dep chain I think | 13:03 |
jamespage | as xenial does not show the same issues | 13:03 |
LuisM | patdk-wk: config-4.4.0-21-generic | 13:05 |
LuisM | rox | 13:05 |
LuisM | I was doing a for to change state of memory to "online" | 13:05 |
LuisM | in /sys/devices/system/memory/memory/*/state | 13:06 |
LuisM | ;) | 13:06 |
jamespage | coreycb, urgh | 13:07 |
jamespage | https://launchpadlibrarian.net/281726807/buildlog_ubuntu-yakkety-amd64.python-oslo.db_4.12.0-1_BUILDING.txt.gz | 13:07 |
jamespage | our switch in default dialect in sqlalchemy bites us at long last | 13:07 |
coreycb | jamespage, I think I've patched tests like that before because of the switch | 13:08 |
coreycb | jamespage, but I wonder if we should revisit that switch | 13:09 |
coreycb | jamespage, python-mysqldb now uses the mysqlclient fork so it has py3 support | 13:12 |
jamespage | grag | 13:12 |
coreycb | jamespage, anyway too late for this cycle | 13:14 |
coreycb | jamespage, for the openssl errors, I wonder if this has something to do with it: https://github.com/openstack/requirements/blob/master/upper-constraints.txt#L95 | 13:28 |
coreycb | jamespage, we're at 1.5 in yakkety-proposed | 13:28 |
=== JanC_ is now known as JanC | ||
jamespage | coreycb, I think so - I see google hits on other projects seeing similar issues | 13:35 |
jamespage | that said that was an autogenerated upper-constraints that fixed that at 1.4 | 13:35 |
zioproto | jamespage: I confirm that removing SKIP_UPDATES and SKIP_PROPOSED lines fixed the problem | 13:51 |
zioproto | jamespage, coreycb: any change to fix the documentation on this page ? https://wiki.ubuntu.com/SimpleSbuild | 13:51 |
coreycb | zioproto, personally I'd be hesitant to update that one since it's not mine, is there a bug? | 13:53 |
zioproto | jamespage: can you recap the bug about python-dh ? | 13:53 |
jamespage | zioproto, I've updated the wiki | 13:54 |
zioproto | coreycb: you replied by email that is metter to push multiple commits when contributing, while james suggested git commit amend and force push :) Both are fine to me. Just agree on one that makes everyone happy :) | 13:54 |
jamespage | made it more opininated 'might -> will' | 13:54 |
jamespage | and added a comment why | 13:54 |
jamespage | zioproto, I prefer a single commit per change - if you need to tweak a commit, its still a single commit | 13:55 |
jamespage | not multiple imho | 13:55 |
coreycb | zioproto, for little changes like a typo I'm ok with a new commit, but ideally a clean history is what we want to see | 13:56 |
zioproto | I see the point. So if I am working with a Yakketi schroot I can safely put SKIP_UPDATES=1. But trusty is old so I need the updates | 13:57 |
zioproto | OK guys, I hope my contributions will get cleaner with time | 13:57 |
jamespage | zioproto, tbh I run all my chroorts with updates enabled | 14:00 |
jamespage | so that at the point of release its a no-op to switch | 14:00 |
blut | how can i setup nis with a preseeded configuration? | 14:00 |
Braven | what does it mean if after selecting the kernel in grub. The screen goes black | 14:01 |
zioproto | talk to you later, I have another meeting, 1 hour away from keyboard :) thanks for the help! | 14:01 |
=== seg_ is now known as seg | ||
LuisM | patdk-wk: your udev rule didn't work :( | 14:10 |
patdk-wk | oh? maybe something more has changed in 4.4 kernel | 14:15 |
ddellav | can i get an archive admin promote python-yaql and python-monascaclient to main please? There were approved MIR's but no promotion: https://bugs.launchpad.net/ubuntu/+source/python-yaql/+bug/1586069 https://bugs.launchpad.net/ubuntu/+source/python-monascaclient/+bug/1590836 | 14:16 |
ubottu | Launchpad bug 1586069 in python-yaql (Ubuntu) "[MIR] python-yaql" [Undecided,Fix released] | 14:16 |
ubottu | Launchpad bug 1590836 in python-monascaclient (Ubuntu) "[MIR] python-monascaclient" [Undecided,Fix released] | 14:16 |
ddellav | they are needed for heat | 14:17 |
jamespage | coreycb, oslo.db fixed up | 14:21 |
jamespage | ddellav, that might be better asked in -devel | 14:22 |
=== alexisb-afk is now known as alexisb | ||
ddellav | jamespage ack | 14:23 |
jamespage | ddellav, its possible that if the heat version that required those new deps was not in main at the time of promotion, they would have been demoted again | 14:23 |
ddellav | coreycb heh, keystonemiddleware is already at 4.9, it's just unreleased | 14:24 |
ddellav | jamespage that's what coreycb was saying, yea | 14:24 |
coreycb | ddellav, that makes it easier, I wonder if it didn't get release for a reason though. jamespage may know. | 14:27 |
jamespage | ddellav, when I looked last week depends in experimental where not quite up-to-date | 14:28 |
ddellav | coreycb looks like it was updated 5 days ago | 14:28 |
jamespage | I pushed as much of my local work as I could | 14:28 |
jamespage | as everything else would catchup | 14:28 |
ddellav | jamespage same deal with keystoneauth1? | 14:31 |
jamespage | yah | 14:31 |
jamespage | gah I need to fixup python-hacking still | 14:31 |
jamespage | yikes | 14:31 |
coreycb | ddellav, can you note that in bug 1579670 that you tested designate-dashboard and it's working in yakkety? be good to let them know. | 14:38 |
ubottu | bug 1579670 in designate-dashboard (Ubuntu) "Designate dashboard 2.0.0-1 throws exception in Xenial" [Undecided,New] https://launchpad.net/bugs/1579670 | 14:38 |
ddellav | coreycb done | 14:39 |
jamespage | coreycb, ddellav: python-hacking munged to work again | 14:53 |
jamespage | its not submittable upstream atm | 14:53 |
coreycb | jamespage, that's good, so we're sorted out for yakkety? | 14:54 |
jamespage | coreycb, yah | 14:54 |
jamespage | basically its a 'skipall' | 14:54 |
jamespage | but tbh I'm less worried about hacking tests being skipped in distro | 14:54 |
coreycb | jamespage, I agree | 14:55 |
xnox | jamespage, coreycb: i have patches for ceilometer test suite. Am I allowed to just upload it and commit/push to https://git.launchpad.net/~ubuntu-server-dev/ubuntu/+source/ceilometer ? or should I raise a merge proposal against it? | 15:01 |
xnox | what's the drive-by contribution workflow for server team? | 15:01 |
jamespage | xnox, oh I think coreycb was working those as well? | 15:02 |
xnox | heh | 15:02 |
coreycb | xnox, contributions are always welcome :) | 15:03 |
coreycb | xnox, was it etc updates or something else? | 15:03 |
xnox | coreycb, what is "etc" ? | 15:03 |
xnox | i'm looking at ADT test suite failure for s390x, which looks like missing depends in the debian/tests/control and/or packages themself. | 15:04 |
xnox | e.g. python-libvirt is now needed to start the services and test that they are available. | 15:04 |
xnox | however not sure how come things succeed on other arches for that test, as i'm failing to see how that would ever work =) | 15:04 |
xnox | i'll send a merge proposal to discuss =) | 15:04 |
coreycb | xnox, there was just some config file updates that I was working on. ah, ADT fixes would be a welcome addition, I wasn't working on that. | 15:05 |
coreycb | xnox, thanks | 15:05 |
zioproto | coreycb, jamespage : I am going offline, let me know with launchpad comments if I can improve my patch. Thank you | 15:23 |
xnox | coreycb, ceilometer-agent-ipmi fails to start and in the journal it tells me that EmptyPollstersList: No valid pollsters can be loaded with the startup parameters polling-namespaces and pollster-list. | 15:26 |
xnox | is there a way to have some "dummy" ipmi pollster-list? or what is missing to get that going? | 15:26 |
xnox | note that i'm inside a lxd container for adt tests on s390x, rather than bare metal =/ | 15:26 |
* xnox ponders if ipmi is a thing on s390x | 15:26 | |
xnox | looking at https://code.launchpad.net/~bladernr/checkbox/no-ipmi-s390x/+merge/291946 i'm guessing no | 15:27 |
xnox | coreycb, i am pondering of how useful ceilometer-agent-ipmi is on s390x | 15:39 |
xnox | given the lack of ipmi on that platform. | 15:39 |
stevenm | is there need for putting 'NEED_IDMAPD=yes' in /etc/default/nfs-common now a redundant one (in 14.04 onwards) as I can see with just nfs-kernel-server installed - that rpc.imapd is running | 16:58 |
stevenm | (sorry for repost, i realised this chan was a better bet) | 16:58 |
=== iberezovskiy_ is now known as iberezovskiy_off | ||
coreycb | xnox, I agree if s390x doesn't have ipmi support then ceilometer-agent-ipmi isn't going to be of much use | 17:03 |
xnox | coreycb, i'll skip testing that in the autopkgtest, but keep everything else in place for now. | 17:04 |
coreycb | xnox, ok sounds good | 17:04 |
coreycb | ddellav, are you done with the oslo's too? | 17:06 |
ddellav | coreycb: not yet, I'll post with packages for your review in a bit | 17:07 |
coreycb | ddellav, ok I can take some and work on them this afternoon if you haven't started on them. let me know. | 17:09 |
cscf | Anyone know of a tool to multithread encryption? Trying to zfs send over ssh and ssh's encryption is a bottleneck | 17:16 |
RoyK | cscf: known problem - using mbuffer will help, but won't give you encryption | 17:24 |
RoyK | cscf: so I guess it all depends on whether you're on a provate network or are using zfs send over the open internet | 17:25 |
cscf | RoyK, I thought surely there must be a tool that AES-encrypts in a pipeline in parallel, like pigz for gzip | 17:25 |
cscf | RoyK, Internal network but a big one, encryption preferred | 17:25 |
RoyK | cscf: ok - not sure - sorry | 17:26 |
coreycb | ddellav, did you do neutron-lib? if not I'll do that. neutron needs it. | 17:33 |
coreycb | ddellav, neutron-lib uploaded, that should at least fix some of the current neutron failures | 18:21 |
xnox | coreycb, jamespage: please review https://code.launchpad.net/~xnox/ubuntu/+source/ceilometer/+git/ceilometer/+merge/304406 | 18:34 |
xnox | darn, i need to rebase i think | 18:35 |
coreycb | xnox, sorry, yeah I pushed some changes under you | 18:36 |
xnox | coreycb, repushing, lp should regenerate in a second or two | 18:37 |
xnox | coreycb, looking good now. | 18:41 |
coreycb | xnox, ok looking | 18:41 |
coreycb | xnox, pushed, thanks! ok if that doesn't get uploaded for another week? we are expecting upstream to release beta 3 this week or early next week. | 18:48 |
xnox | coreycb, that's all good. | 18:50 |
xnox | coreycb, hopefully next upload with actually migrate to yakkety from -proposed. | 18:50 |
xnox | coreycb, at the moment yakkety release is still at 6.0.0 and b2 is stuck in yakkety-proposed, but my patches should fix that. | 18:51 |
coreycb | xnox, let me upload then so we can make sure it migrates from proposed. it's a little tricky ,but manageable, because we've been pushing updates to keep up with upstream tip. | 18:54 |
=== JanC is now known as Guest64850 | ||
=== JanC_ is now known as JanC | ||
ajikny | i have an intersting problem to solve. I just upgraded the server form ubuntu 12.04 to 14.04. As soon as the upgrade got over, which went smooth, | 21:53 |
ajikny | After the reboot the keyboard and mouse stopped working. | 21:53 |
ajikny | After the reboot the keyboard and *network stopped working | 21:54 |
ajikny | so i neither i can ssh into the server nor can i type anything. | 21:54 |
ajikny | What can i do?? | 21:54 |
JanC | ajikny: you have physical access to the server to confirm it actually booted? | 22:16 |
ajikny | yess | 22:17 |
ajikny | i was in front of it. i see the tty | 22:17 |
ajikny | i have livebooted and kb works in the live boot and uptil grub | 22:18 |
ajikny | i have got syslogs. if you are interested | 22:18 |
JanC | I (or others) can look at them maybe | 22:20 |
ajikny | ya sure, ill upload them in a sec and share the link | 22:21 |
ajikny | http://s000.tinyupload.com/?file_id=43836141207735899929 | 22:22 |
sarnold | can you use a site that doesn't serve junk ads? | 22:24 |
ajikny | sorrry | 22:24 |
sarnold | this is polite :) http://paste.ubuntu.com/ | 22:24 |
=== i- is now known as OverACoder | ||
=== OverACoder is now known as i- | ||
JanC | I didn't really see any ads :) | 22:26 |
sarnold | I made the mistake of enabling the site in noscript to try to download the original file | 22:26 |
sarnold | I never saw the file, just junk.. | 22:26 |
ajikny | http://paste.ubuntu.com/23113849/ | 22:28 |
sarnold | ahh much better, thanks ;) | 22:28 |
ajikny | :) | 22:29 |
sarnold | it sure looks like the keyboard disconnects and reconnects every minute | 22:30 |
sarnold | is it a broken KVM switch perhaps? or a broken keyboard? | 22:30 |
JanC | there seem to be a lot of mouse connect/disconnects | 22:30 |
sarnold | broken usb switch? | 22:30 |
ajikny | the keyboard works very normally in live boot and in grub. | 22:31 |
JanC | Sachiru: not sure where you see keyboard disconnects? | 22:31 |
JanC | sarnold: I mean | 22:31 |
sarnold | JanC: `you're right they're all mice; I just assumed it'd be a keyboard on a server and didn't read it all that closely | 22:32 |
JanC | ajikny: is that a an USB keyboard? | 22:33 |
ajikny | yes | 22:33 |
JanC | there seems to be an USB keyboard directly attached, and one attached to or emulated by an USB KVM switch? | 22:37 |
RoyK | s/an/a/ | 22:37 |
RoyK | :P | 22:37 |
ajikny | There is only one that is attached.. | 22:39 |
ajikny | I dont follow, Does kernel accept only the emulated kb?? | 22:41 |
JanC | it should not | 22:42 |
JanC | things can break when a broken keyboard (real or emulated) is attached though | 22:43 |
JanC | still wouldn't explain networking, I guess | 22:43 |
ajikny | Yes, they must be unrelated. | 22:43 |
ajikny | =https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1313497 Can this be relevent?? | 22:45 |
ubottu | Launchpad bug 1313497 in linux (Ubuntu) "USB keyboard unresponsive on fresh install" [Undecided,Invalid] | 22:45 |
ajikny | :) | 22:46 |
JetNet | hi | 22:51 |
JanC | ajikny: did the networking need special drivers that are not part of the default Ubuntu kernel? | 22:55 |
JanC | or were not | 22:55 |
JanC | also “Aug 30 13:20:01 mail mdadm[1471]: DegradedArray event detected on md device /dev/md/1” | 22:57 |
ajikny | JanC: no, just confirmed with previous sys admin. on 12.04 The nework ran on the default kernel | 23:01 |
ajikny | This error must refer to the one disk of raid 1 which is wearing out. . The machine has a softwware raid | 23:01 |
JanC | “Aug 30 13:19:58 mail kernel: [ 21.372184] init: failsafe main process (1071) killed by TERM signal” | 23:02 |
JanC | not sure where that comes from | 23:03 |
danny_ | What NAS protocol do you guys recommend for being used as additional storage to one ubuntu computer? | 23:03 |
JanC | ajikny: for the keyboard I would suggest trying to use one keyboard/KVM at a time & maybe try a different keyboard too | 23:04 |
JanC | and try different USB ports & disconnect all other USB devices | 23:04 |
JanC | just to make sure it's not a hardware issue | 23:05 |
JanC | maybe try PS/2 :) | 23:05 |
JanC | ajikny: did you see my suggestion? | 23:10 |
ajikny | JanC: I have connected on all ports and have different keyboards, | 23:10 |
ajikny | yes.. | 23:10 |
ajikny | i lost connection for a while | 23:10 |
JanC | no network card detected in the kernel logs from what I can tell | 23:12 |
JanC | danny_: I would use NFS, but that's mostly because Samba needs special options to be able to use filenames that contain "?" and such over SMB/CIFS :) | 23:15 |
danny_ | JanC, Ok, I was planning on using NFS. Now my only problem is getting write permissions on my pc | 23:17 |
ajikny | okay.. That is grave | 23:20 |
ajikny | JanC: if in the live boot i mount the drive and bind /dev /dev/pts/ /proc/ /sys /run ... followed by chroot mount point will my commands run as if i was on host system | 23:21 |
sarnold | danny_: 'easiest' is to ensure that youv'e got the same userid numbers on both systems | 23:23 |
danny_ | sarnold, Sorry I'm new to this, what exactly does that mean? Is userid different than username? | 23:25 |
sarnold | danny_: yes; unix assigns every user a number for their userid, a number for groupid, and numbers for group memberships | 23:26 |
sarnold | danny_: the names often suffice but NFS uses the numbers; it expects them to be common to all clients and servers | 23:26 |
sarnold | danny_: in larger installments people may use NIS or NIS+ or LDAP or something similar ot synchronize user accounts across all their machines | 23:26 |
sarnold | with just two machines it's probably easier to create users with the correct account nubmers from the start, or manually modify them as needed | 23:27 |
danny_ | sarnold, ok, I gotta eat then I'mma learn all of this | 23:28 |
trippeh_ | I just used sshfs for this for a while. Wasn't too bad actually! | 23:28 |
trippeh_ | even the torrent client was happy haha. | 23:29 |
sarnold | trippeh_: ha I figured you'd use xprtrdma :) | 23:30 |
trippeh_ | hehe yeah sshfs didnt scale too well to 10G | 23:31 |
trippeh_ | topped out at a hair over 2Gbps | 23:31 |
sarnold | trippeh_: patdk's got some ssh patches he uses for high-throughput ssh.. | 23:32 |
sarnold | https://www.psc.edu/index.php/hpn-ssh | 23:32 |
trippeh_ | yeah, they arent too effective for local stuff, unless you hit cpu limits on the crypto parts | 23:34 |
trippeh_ | nowadays just use aesni ;) | 23:35 |
JanC | some NAS devices have settings to allow all NFS clients equal access (haven't looked at how exactly they implement this) | 23:37 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!