--- Log opened vr jun 20 00:00:14 2014 00:14 -!- adahms (Andrew Dahms): has joined #vdsm 01:10 -!- nsoffer: has quit [Ping timeout: 240 seconds] 01:50 -!- moolit: has quit [Quit: Leaving.] 01:52 -!- mpolednik: has quit [Ping timeout: 264 seconds] 03:13 -!- bala (purple): has joined #vdsm 04:01 -!- mliu (purple): has joined #vdsm 04:01 -!- mliu: has left #vdsm [] 04:05 -!- bala: has quit [Quit: Leaving.] 04:45 -!- mliu1 (purple): has joined #vdsm 04:45 -!- mliu1: has left #vdsm [] 05:04 -!- shaohef: has left #vdsm [] 05:05 -!- shaohef (shaohef): has joined #vdsm 06:32 -!- adahms: has quit [Quit: Leaving] 06:33 -!- adahms (Andrew Dahms): has joined #vdsm 06:54 -!- mliu (purple): has joined #vdsm 06:57 -!- mliu: has quit [Client Quit] 06:58 -!- mliu (purple): has joined #vdsm 06:59 -!- mliu: has left #vdsm [] 07:01 -!- gpadgett: has quit [Ping timeout: 244 seconds] 07:05 -!- hchiramm_ (Humble Chirammal): has joined #vdsm 07:24 -!- xfrancis (Xavi Francisco): has joined #vdsm 07:31 -!- xfrancis: has quit [Quit: Leaving] 07:31 -!- xfrancis (Xavi Francisco): has joined #vdsm 07:33 -!- aravindavk (Aravinda): has joined #vdsm 07:42 -!- bala (purple): has joined #vdsm 08:08 -!- adahms: has quit [Quit: Leaving] 08:18 -!- hchiramm_: has quit [Ping timeout: 255 seconds] 08:25 -!- timothy (Timothy Asir): has joined #vdsm 08:29 -!- aravindavk: has quit [Ping timeout: 245 seconds] 08:29 -!- fromani (Francesco Romani): has joined #vdsm 08:32 -!- pkliczew (Piotr Kliczewski): has joined #vdsm 08:57 -!- aravindavk (Aravinda): has joined #vdsm 09:03 -!- hchiramm_ (Humble Chirammal): has joined #vdsm 09:20 -!- mliu (purple): has joined #vdsm 09:27 -!- mliu: has left #vdsm [] 09:27 -!- mliu (purple): has joined #vdsm 09:27 -!- mliu: has left #vdsm [] 09:53 -!- apuimedo (antoni): has joined #vdsm 10:23 -!- mode/#vdsm: by ChanServ 10:29 < apuimedo> danken: boker tov 10:29 <@danken> apuimedo: bon dia 10:29 <@danken> apuimedo: (slow morning for me) 10:29 < apuimedo> danken: how's 3.5 cooking? 10:31 <@danken> apuimedo: dunno yet; I did not see if msivak/kobi has posted the new mom dependency yet. ioprocess does not look good. 10:31 < apuimedo> danken: anything I can help with (besides ovirt-node)? 10:32 < msivak> danken: mom 0.4.1 should be ready, the patch with the dependency is http://gerrit.ovirt.org/#/c/28911/ 10:32 < msivak> danken: I am waiting for the repositories to get the updated version 10:32 <@danken> apuimedo: nothig at the momemnt, thanks 10:32 < apuimedo> ok 10:33 < apuimedo> danken: did you see that mperina verified the boot time ifup serialization patch? 10:33 <@danken> msivak: cool. you should also get it into our jenkins slave (in case it is not automatic) 10:33 <@danken> apuimedo: yes! but I wanted to chat with you about it 10:33 <@danken> I mean, that was just a surprise hunch of mine. did it completely solve his issue? 10:34 < apuimedo> as I expected, it did 10:34 <@danken> apuimedo: (It's quite sad to repeat the same expression 4 times... not that I wish for a big instrumentation to change that) 10:35 <@danken> msivak: does the mom build have enough karma? if so, where is it stuck? 10:35 < msivak> danken: it is already in F20 10:35 < msivak> danken: I just checked and I am trying to verify 10:35 < apuimedo> well, I can put it in a helped method... 10:36 < apuimedo> *helper 10:36 <@danken> apuimedo: that's exactly what I find over-burdenning 10:36 < msivak> danken: it fixes the getVdsStats issue 10:36 < msivak> danken: there is still a small bug in mom, but it does not affect vdsm 10:37 < apuimedo> danken: well, I guess that objectivizeNetwork could be taught to set ipConfig.async = False when in rollback 10:37 < apuimedo> I'll see how it looks like 10:43 <@danken> apuimedo: thanks, though it's not horribly important 10:43 < apuimedo> danken: the refactor or the fix? 10:44 <@danken> the refactor 10:44 < apuimedo> ok 10:44 <@danken> (the bugfix is important! thanks!) 10:44 < apuimedo> well, it's not easy to reproduce, but for the people that encounter it it's quite damning 10:44 < apuimedo> :P 10:57 < apuimedo> danken: what about this more targeted approach http://gerrit.ovirt.org/#/c/28935/2 11:01 -!- mliu (purple): has joined #vdsm 11:02 -!- mliu: has left #vdsm [] 11:02 -!- timothy: has quit [Ping timeout: 244 seconds] 11:09 <@danken> apuimedo: I find it more elegant! 11:10 < apuimedo> danken: it is. I originally chose the other approach not to impact iproute2 which I would be surprised to see that it would suffer from the same issue 11:10 < apuimedo> but consistency and cleanliness is important ;-) 11:15 -!- timothy (Timothy Asir): has joined #vdsm 11:16 <@danken> apuimedo: nit question: why is the ipv6autoconf=ipv6autoconf reordering? 11:17 < apuimedo> danken: heh, it almost fitted in two lines 11:17 < apuimedo> then I noticed a mistake 11:17 < apuimedo> and it didn't fit anymore 11:18 < apuimedo> I'll fix the reordering 11:22 < apuimedo> danken: done 11:31 < fromani> pkliczew: thanks for your review on http://gerrit.ovirt.org/#/c/28802/ and for the pointer to Bridge.py! 11:32 < fromani> pkliczew: do you mind a few questions about Bridge.py in the next days? 11:32 < pkliczew> fromani: I though it would be useful to use it for verification 11:32 < pkliczew> fromani: happy to answer any questions you will have 11:32 < fromani> pkliczew: awesome, tnx! 11:38 <@danken> apuimedo: there's beauty in alonbl's arg-per-line style: you do not need to think about these nonsense. 11:39 < apuimedo> danken: I only follow that style when programming in verilog 11:39 < apuimedo> xD 11:44 < msivak> apuimedo: verilog? I didn't know you are a hardware guy :) 11:45 < apuimedo> msivak: I used to do some stuff in the end of university, some open cores 11:45 < apuimedo> I probably forgot most of it 11:45 < msivak> apuimedo: interesting.. most EU universities teach vhdl :) 11:46 < apuimedo> msivak: the project was led by a guy from ARM and the course by some guy of the Barcelona Intel reseach lab 11:46 < apuimedo> I guess they had some reason 11:46 < apuimedo> probably just what the arm guy was used to 11:46 < msivak> apuimedo: US companies prefer verilog :) 11:46 < apuimedo> I've never used vhdl so I can't tell which is better 11:46 < msivak> but the code can actually be converted from one to another quite easily 11:47 < msivak> apuimedo: vhdl is more strongly typed and more verbose 11:47 < msivak> apuimedo: like java vs C :) 11:47 < apuimedo> ah, i see 11:47 < apuimedo> I prefer verilog then 11:47 < apuimedo> xD 12:01 -!- timothy: has quit [Ping timeout: 240 seconds] 12:06 < apuimedo> danken: thanks for the merge 12:14 -!- timothy (Timothy Asir): has joined #vdsm 12:30 -!- sbonazzo: has quit [Quit: Leaving.] 12:36 <@danken> msivak: alitke is expected to wake up only later today - could you "own" your suggested followup to http://gerrit.ovirt.org/#/c/28966/ ? 12:44 < apuimedo> danken: extra lines are evil :P 12:46 <@danken> apuimedo: touching unrelated lines is evil! 12:46 < apuimedo> danken: I'll admit to that 12:46 < apuimedo> :P 12:51 < msivak> danken: do you want a separate patchset or should I just add the fix and resubmit this one? 12:52 <@danken> msivak: one would do for me 12:52 <@danken> it's your call. 12:55 < apuimedo> dougsland: could you please verify http://gerrit.ovirt.org/#/c/28879/ 12:55 < apuimedo> so that you don't need to manually unpersist the ifcfg-eth0 anymore? 12:56 -!- evilissimo|afk is now known as evilissimo 13:07 < apuimedo> going for lunch 13:11 -!- apuimedo: has quit [Ping timeout: 244 seconds] 13:47 -!- timothy: has quit [Quit: Leaving] 13:49 -!- AlbertoSilva (AlbertoSilva): has joined #vdsm 14:01 -!- xfrancis: has quit [Ping timeout: 264 seconds] 14:03 -!- mskrivanek_away is now known as mskrivanek 14:05 <@danken> msivak: thanks for the new patch. I hate to nag about petty things, but the commit message is left a bit out-of-date 14:05 < msivak> danken: I updated it 14:05 <@danken> "Fix by using dict.get() instead of direct indexing" 14:05 < msivak> ah 14:05 < msivak> that part 14:05 <@danken> (I prefer your solution, of using "in") 14:06 <@danken> msivak: so in case you care, take the opportunity to reword! 14:07 < msivak> I just did 14:27 <@danken> thanks 14:28 < fromani> danken: maybe I'm missing something, but I'm not sure the AdvancedStatsFunction is working as advertised with window != 2 14:29 <@danken> argg, I've merged your 'fist checking' msivak. 14:30 <@danken> ;-) 14:30 <@danken> fromani: whatdy'a mean? it does not maintain window samples ? 14:31 < fromani> danken: nevermind, I was just mislead by a comment. My fault. 14:45 -!- mskrivanek is now known as mskrivanek_away 14:56 -!- aravindavk: has quit [Ping timeout: 240 seconds] 15:07 < alitke> danken / msivak: mom-0.4.1-1 is now an update in Fedora 20 15:07 < alitke> When are we ready for a patch to require it in the vdsm spec file? 15:07 < alitke> I want to ensure that I disrupt as few people as possible with mom integration issues. 15:13 < msivak> alitke: see http://gerrit.ovirt.org/#/c/28977/ - happens only on F20 with specific libvirt (which I happen to have..) so we might want to merge it as well 15:14 < msivak> alitke: about the spec file patch - http://gerrit.ovirt.org/#/c/28911/ 15:14 < alitke> msivak, That's not going to make 3.5 since I just did a release 15:16 < alitke> msivak, there is a longer delay for el6 and I still need to package for el7. 15:16 < alitke> which is a trivial merge, push, build, update cycle 15:17 < msivak> el7? well I own the errata so tell me when you have it :) 15:17 < alitke> I mean epel7 15:17 < msivak> ah, ok 15:18 < alitke> I guess that matters to us since it will support centOS 15:21 < msivak> eventually, I do not think there is centos 7 atm 15:23 < ewoud> well, epel7 is in beta and centos 7 is being prepared but doesn't have all packages yet 15:31 < alitke> ok 15:32 -!- gpadgett (Greg Padgett): has joined #vdsm 15:37 -!- apuimedo (antoni): has joined #vdsm 15:41 -!- apahim: has quit [Read error: Connection reset by peer] 15:44 -!- bala: has quit [Ping timeout: 244 seconds] 15:56 < fromani> just spotted while testing on RHEL6.5: http://pastebin.test.redhat.com/216723 15:56 < fromani> msivak: is that a known issue? ^^ 16:01 < msivak> fromani: no, but I have a libvirt guy right next to me, so it is now :) 16:02 < fromani> msivak: I'm using libvirt-0.10.2-29.el6_5.7.x86_64 16:02 < fromani> AFAIK is the last one available in RHEL 6.5 16:03 < fromani> and qemu-kvm-rhev-0.12.1.2-2.414.el6.x86_64 16:03 < msivak> fromani: might be, but libvirt api supports that function, they just never filled the body.. well till some time ago :) 16:03 < apuimedo> xD 16:03 < msivak> libvirt bug :) 16:28 < dougsland> danken, ready to go the toni's patch: http://gerrit.ovirt.org/#/c/28879/ 16:28 < dougsland> apuimedo, ^ 16:29 < apuimedo> dougsland: thank you very much! 16:34 <@danken> dougsland: apuimedo: taken, thanks folks! 16:34 < apuimedo> ;-) 16:34 <@danken> could you now pet the bugs ? 16:34 < apuimedo> you're welcome 16:35 < apuimedo> danken: bug 16:35 < apuimedo> only one 16:35 < apuimedo> this doesn't fix the other one 16:35 < dougsland> apuimedo, danken sure, np. 16:35 < dougsland> apuimedo, yes, the host currently cannot reboot 16:36 < apuimedo> danken: I didn't do anything, but https://bugzilla.redhat.com/show_bug.cgi?id=1104035 just went to modified 16:36 < apuimedo> anything else needed? 16:47 -!- sbonazzo: has quit [Quit: Leaving.] 16:50 <@danken> apuimedo: wasn't there a second node issue? 16:51 < apuimedo> yes 16:51 < apuimedo> on which I'm working 16:51 < apuimedo> :-) 16:51 -!- bazulay (purple): has joined #vdsm 16:52 <@danken> pkliczew: thanks for the second round of the lastClient work. I'm yet to read it. 16:53 <@danken> pkliczew: sorry to be nagging, but when API changes, the commit message should be firm. 16:53 < pkliczew> danken: you are welcome! 16:53 < pkliczew> danken: Sure, I understand 16:54 < pkliczew> danken: Please reply to Nir's question from patch set #3 16:55 <@danken> could you give me the direct link 16:57 <@danken> pkliczew: ?^^^ 16:57 < pkliczew> danken: sure 16:58 < pkliczew> http://gerrit.ovirt.org/#/c/28817/3..4/vdsm/rpc/BindingXMLRPC.py 16:58 < pkliczew> line 46 17:02 <@danken> pkliczew: I'll take a look 17:02 < pkliczew> danken: Great, Thank you! 17:06 -!- nsoffer (Nir Soffer): has joined #vdsm 17:09 -!- pkliczew: has quit [Quit: Leaving] 17:16 -!- bala (purple): has joined #vdsm 17:26 -!- bala: has quit [Ping timeout: 240 seconds] 17:27 -!- bala (purple): has joined #vdsm 17:42 < msivak> danken: a question, do we use direct dom manipulation to set data when libvirt lacks some api or is it not recommended? 17:42 < apuimedo> ? 17:42 < apuimedo> can you put an example? 17:42 < msivak> apuimedo: libvirt domxml supports iotune for disks 17:42 < msivak> apuimedo: but there is no api to change the value 17:43 < apuimedo> isn't the api the xml update? 17:43 < msivak> apuimedo: there are direct methods for some calls 17:44 < apuimedo> but for some, like iotune there is not, right? 17:44 < msivak> but this one is missing, so I suppose I would have to update the xml 17:45 <@danken> msivak: I do not understand - what is your alternative to using libvirt API? 17:46 < msivak> danken: well read, update and save the xml for the domain 17:46 < apuimedo> msivak: if the question is if we can call the api to update the xml 17:46 < msivak> I guess that is also the api 17:46 < apuimedo> we already do that for some fractions 17:46 < apuimedo> like for updating the vnic part of the xml 17:47 < msivak> ah 17:50 <@danken> msivak: there's no api to change domxml. only to update a specific device 17:50 -!- keds: has quit [] 17:51 < msivak> danken: a device is enough 17:52 <@danken> msivak: double check that it actually work. apuimedo can tell about a silent failure to act upon the xml 17:54 < apuimedo> danken: indeed 17:54 < apuimedo> :D 17:59 < msivak> I will, it will take me some time to actually write the code :) 18:01 < msivak> but I just found out that I know the new Brno libvirt team lead :) so I have a stick (and I know the Brno guys quite well too so a carrot approach is possible as well) 18:01 < msivak> we found a libvirt bug today you might be interested in 18:02 < msivak> if you have libvirt with systemd support (F20 at least) and restart the libvirt daemon (rpm upgrade or something), it will forget about cgroups for running VMs 18:03 < msivak> anyway, have a nice weekend, see you next week 18:04 -!- evilissimo is now known as evilissimo|afk 18:07 -!- fromani: has quit [Quit: Leaving] 18:08 < apuimedo> msivak: have a nice weekend 18:08 < apuimedo> ! 18:21 -!- sbonazzo: has quit [Client Quit] 18:34 -!- sbonazzo: has quit [Quit: Leaving.] 18:40 -!- bala: has quit [Quit: Leaving.] 19:01 -!- bala (purple): has joined #vdsm 19:10 -!- bala: has quit [Quit: Leaving.] 20:16 -!- fabiand (Fabian Deutsch): has joined #vdsm 20:16 < fabiand> dougsland, hey 20:16 < fabiand> danken, dougsland -- I've pushed two patches 20:16 < fabiand> One is new - basically some mechanism to detect image changes 20:16 < fabiand> what we've spoken about yeterday 20:17 < fabiand> please review: http://gerrit.ovirt.org/29004 http://gerrit.ovirt.org/29005 20:17 -!- fabiand: has quit [Client Quit] 20:44 -!- danken: has quit [Quit: Leaving.] 21:27 -!- mode/#vdsm: by ChanServ 21:31 -!- apahim: has quit [Quit: Leaving] 22:06 -!- orc_emac_: has quit [Ping timeout: 252 seconds] 22:06 -!- orc_emac ("First Av orc_emac"): has joined #vdsm 22:07 -!- orc_emac: has quit [Changing host] 22:07 -!- orc_emac ("First Av orc_emac"): has joined #vdsm 22:25 < apuimedo> danken: found another bug in the ovirt-node 22:25 < apuimedo> will send a patch straight away 22:31 -!- moolit (purple): has joined #vdsm 22:35 < apuimedo> thre 22:35 < apuimedo> *there, pushed 22:41 < apuimedo> nsoffer: doing patches late, huh? 22:41 < apuimedo> and on a weekend! 22:41 < nsoffer> apuimedo, yea :-) 22:41 -!- danken: has quit [Quit: Leaving.] 22:42 < apuimedo> nsoffer: our wives will abandon us 22:42 < apuimedo> xD 22:42 < nsoffer> apuimedo, you are invited to review if you are bored 22:42 < nsoffer> apuimedo, mine sleeping 22:42 < apuimedo> nsoffer: I'm very entertained sending patches to ovirt-node 22:42 < apuimedo> ah, mine reading 22:42 < apuimedo> first bug fixes 22:42 < apuimedo> then refactoring 22:42 < apuimedo> and a feature/fix 22:43 < nsoffer> apuimedo, no fix yet for ONBOOT issue - right? 22:43 < apuimedo> nsoffer: heh 22:43 < apuimedo> no branching no fix 22:43 < apuimedo> :P 22:43 < nsoffer> master? 22:44 < apuimedo> master will have a nicer fix 22:44 < nsoffer> I kind of tired loosing my machines after installing vdsm 22:44 < apuimedo> being that 3.6 is el7 22:44 < apuimedo> and I can send the initscripts to suck it 22:44 < apuimedo> sorry, I mean the sysV scripts 22:44 < nsoffer> you said that I can set some config value to fix this - I forget what was that 22:45 < apuimedo> ONBOOT 22:45 < apuimedo> if you put it to /var/lib/vdsm/persistence/nets/ovirtmgmt 22:45 < nsoffer> setting ONBOOT on ifcfg-x does not help 22:45 < apuimedo> "ONBOOT": true 22:45 < nsoffer> can you show me an example file? 22:46 < apuimedo> nsoffer: can I ssh into your machine? 23:30 -!- moolit: has quit [Quit: Leaving.] 23:47 -!- #vdsm,#centos fsimonce: has quit [Quit: Coyote finally caught me] --- Log closed za jun 21 00:00:16 2014