--- Log opened wo jun 24 00:00:58 2015 00:10 -!- gpadgett: has quit [Quit: Leaving] 00:33 -!- timothy_ (Timothy Asir): has joined #vdsm 00:36 -!- adahms (Andrew Dahms): has joined #vdsm 01:06 -!- nsoffer: has quit [Ping timeout: 276 seconds] 02:31 -!- timothy_: has quit [Ping timeout: 250 seconds] 04:09 -!- #vdsm dougsland: has quit [Ping timeout: 250 seconds] 04:32 -!- bala (purple): has joined #vdsm 04:57 -!- apuimedo: has quit [Ping timeout: 264 seconds] 05:14 -!- shubhendu (Shubhendu Ram Tripathi): has joined #vdsm 05:22 -!- bala: has quit [Quit: Leaving.] 05:22 -!- bala (purple): has joined #vdsm 05:24 -!- bala: has quit [Read error: No route to host] 05:26 -!- bala (purple): has joined #vdsm 05:31 -!- bala: has quit [Read error: Connection reset by peer] 06:03 -!- timothy (Timothy Asir): has joined #vdsm 06:03 -!- timothy is now known as Guest36150 06:13 -!- vered (Vered Volansky): has joined #vdsm 06:22 -!- dyasny (Dan Yasny): has joined #vdsm 06:53 -!- Guest36150: has quit [Ping timeout: 248 seconds] 07:07 -!- #vdsm dyasny: has quit [Ping timeout: 246 seconds] 07:14 -!- nsoffer (Nir Soffer): has joined #vdsm 07:23 -!- ndarshan (Darshan n): has joined #vdsm 07:26 -!- Madhu_ (Madhu_): has joined #vdsm 07:43 -!- ishaby (Idan Shaby): has joined #vdsm 07:55 -!- sbonazzo (purple): has joined #vdsm 08:05 -!- nsoffer: has quit [Ping timeout: 256 seconds] 08:07 -!- mskrivanek_away is now known as mskrivanek 08:09 -!- adahms: has quit [Read error: Connection reset by peer] 08:19 -!- ibarkan: has quit [Ping timeout: 246 seconds] 08:24 -!- Guest36150 (Timothy Asir): has joined #vdsm 08:34 -!- pkliczew (Piotr Kliczewski): has joined #vdsm 08:42 -!- mmirecki (Marcin Mirecki): has joined #vdsm 08:43 -!- fabiand (Fabian Deutsch): has joined #vdsm 08:51 -!- shubhendu: has quit [Ping timeout: 246 seconds] 08:56 -!- nsoffer (Nir Soffer): has joined #vdsm 09:00 -!- Guest36150: has quit [Ping timeout: 256 seconds] 09:05 -!- timothy (Tim): has joined #vdsm 09:06 -!- timothy is now known as Guest25860 09:15 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 09:15 -!- ibarkan (purple): has joined #vdsm 09:17 -!- fromani (Francesco Romani): has joined #vdsm 09:22 -!- timothy_ (Timothy Asir): has joined #vdsm 09:31 -!- derez_ (Daniel Erez): has joined #vdsm 09:32 -!- ykaplan (Yeela Kaplan): has joined #vdsm 09:38 -!- tjeyasin (Timothy Asir): has joined #vdsm 09:40 -!- acanan (Aharon Canan): has joined #vdsm 09:42 -!- ibarkan: has left #vdsm [] 09:43 -!- phoracek (phoracek): has joined #vdsm 09:55 < fabiand> Hey 09:56 < fabiand> Why is the unified persistence configuration actually in /var and not in /etc? 10:01 < evilissimo1> because it's runtime data and not really application configuration data 10:01 < evilissimo1> at least that's how I would argue here 10:03 -!- nsoffer: has quit [Ping timeout: 244 seconds] 10:04 -!- ybronhei (purple): has joined #vdsm 10:06 -!- ishaby: has quit [Ping timeout: 252 seconds] 10:07 -!- bazulay (purple): has joined #vdsm 10:21 -!- ishaby (Idan Shaby): has joined #vdsm 10:41 < fabiand> evilissimo1, mh 10:41 < fabiand> evilissimo1, but isn't it defining the network configuration? 10:42 < fabiand> I mean: After all it "replaces" the ifcfg files, and they are located in /etc - and: They define the system wider configuration of the network interfaces. 10:42 < fabiand> … indirectly 10:54 < evilissimo1> Well it's saving the configuration received on runtime from the engine, and remembers it so that it will know the correct configuration after a reboot 10:55 < evilissimo1> and but anyway that's rather a question for danken or someone from the networking team, I just tried to make sense of it :D 10:55 -!- nsoffer (Nir Soffer): has joined #vdsm 10:56 -!- phoracek: has quit [Ping timeout: 244 seconds] 10:57 < fabiand> evilissimo1, :) 10:57 < fabiand> yes. 11:00 -!- bala (purple): has joined #vdsm 11:01 -!- mpolednik (mpolednik): has joined #vdsm 11:03 -!- timothy_: has quit [Ping timeout: 244 seconds] 11:04 -!- tjeyasin: has quit [Ping timeout: 256 seconds] 11:12 -!- phoracek (phoracek): has joined #vdsm 11:13 -!- ishaby: has quit [Ping timeout: 252 seconds] 11:13 -!- mmirecki is now known as mmirecki-afk 11:15 -!- tjeyasin (Timothy Asir): has joined #vdsm 11:16 -!- timothy_ (Timothy Asir): has joined #vdsm 11:19 -!- bala: has quit [Ping timeout: 244 seconds] 11:26 -!- ishaby (Idan Shaby): has joined #vdsm 11:36 -!- mskrivanek is now known as mskrivanek_away 11:36 -!- mskrivanek_away is now known as mskrivanek 11:37 -!- bala (purple): has joined #vdsm 11:46 -!- mmirecki-afk is now known as mmirecki 11:49 -!- bala: has quit [Ping timeout: 272 seconds] 11:51 -!- timothy_: has quit [Ping timeout: 250 seconds] 11:51 -!- tjeyasin: has quit [Ping timeout: 244 seconds] 11:57 -!- sshnaidm: has quit [Ping timeout: 276 seconds] 11:58 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 11:59 -!- mode/#vdsm: by ChanServ 11:59 -!- danken (purple): has joined #vdsm 12:00 -!- phoracek: has quit [Ping timeout: 256 seconds] 12:03 -!- tjeyasin (Timothy Asir): has joined #vdsm 12:03 -!- timothy_ (Timothy Asir): has joined #vdsm 12:03 -!- bala (purple): has joined #vdsm 12:08 -!- mpolednik1 (mpolednik): has joined #vdsm 12:08 -!- derez_: has quit [Quit: Leaving] 12:11 -!- mpolednik: has quit [Ping timeout: 256 seconds] 12:12 < fabiand> danken, hey 12:12 < fabiand> danken, I'm wondering why the unfied persistence files are in /var and not in /etc 12:13 -!- phoracek (phoracek): has joined #vdsm 12:13 < fabiand> To me they look like configuration files which are host specific - and are required to bring up the system 12:15 -!- mpolednik1: has quit [Quit: WeeChat 1.2] 12:16 -!- mpolednik (mpolednik): has joined #vdsm 12:19 -!- acanan: has quit [Ping timeout: 250 seconds] 12:25 -!- tjeyasin: has quit [Ping timeout: 264 seconds] 12:25 -!- timothy_: has quit [Ping timeout: 264 seconds] 12:25 -!- Guest25860 is now known as _tim_07 12:34 -!- acanan (Aharon Canan): has joined #vdsm 12:35 -!- dougsland (Douglas): has joined #vdsm 12:36 -!- tjeyasin (Timothy Asir): has joined #vdsm 12:36 -!- timothy_ (Timothy Asir): has joined #vdsm 12:50 -!- sshnaidm: has quit [Ping timeout: 276 seconds] 12:50 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 13:08 < mpolednik> danken: ibarkan2: https://gerrit.ovirt.org/#/c/42793/1 ? 13:08 -!- ishaby: has quit [Quit: Leaving] 13:08 -!- ishaby (Idan Shaby): has joined #vdsm 13:14 < mpolednik> fromani: danken: https://gerrit.ovirt.org/#/c/42802/ ? (did pep8 explode?) 13:19 <@danken> fabiand: /etc should have only stuff edited by humans 13:20 <@danken> mpolednik: huh? is undoing a recent fromani change. 13:20 <@danken> maybe we have an issue with pep8 versions? 13:20 <@danken> mpolednik: which version are you using? 13:28 < mpolednik> python-pep8-1.4.6-3.el7.noarch 13:32 < fabiand> danken, that is a good differentiation 13:32 < fabiand> danken, I was thinking about the container case 13:32 < mpolednik> or is it building fine or others? it is latest master 13:33 < fromani> mpolednik: danken funny indeed, I did that change _exactly_ and _only_ because pep8 complained 13:34 < mpolednik> fromani: which version is it for you? 13:34 < fromani> mpolednik: my 1.5.7 was happy, on CI (can't remember which slave) there it was some complaints 13:38 < amarchuk> danken, ping 13:40 < fromani> danken: hi! please share your thoughts about https://gerrit.ovirt.org/#/c/42687/ - NOT urgent 13:43 -!- sshnaidm: has quit [Remote host closed the connection] 13:45 <@danken> amarchuk: ? 13:46 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 13:47 < amarchuk> danken, can you please check out https://gerrit.ovirt.org/#/c/42573/ from one side it is true that there are more problems with the hook since it hardcodes virtio. from other side the current version is not working and when those changes applied it would work for virotio. I think it is possible to get it accesspted and then we should decide what to do with virtio hardcode and have another patch fixing that. what you think? 13:48 < mpolednik> fromani: so, 1.5.7 works fine, 1.4.6 fails 13:48 < mpolednik> wow 13:48 < mpolednik> 1.5.6 works fine also 13:48 < amarchuk> we are going to become a user of that hook as we really need to have local storage for ci servers, so any necessary patches will follow. right now I have no reliable idea on how to guess the drive type 13:52 -!- bala: has quit [Ping timeout: 264 seconds] 13:53 < mpolednik> fromani: danken: with those patches both pep8s (1.4.6, 1.5.7) pass for me, with 1.4.6 current master fails :/ 13:54 <@danken> amarchuk: I don't quite follow your question... 13:54 < fromani> mpolednik: let's comply to the wishes of our CI overlords :) 13:54 <@danken> I am sure fromani shares you willingness to fix this old hook 13:55 < mpolednik> fromani: and right now, you can get either version on CI :P 13:55 < amarchuk> danken, I think we will fix that it is just imho two separate patches 13:55 < amarchuk> pushing this will make already some use cases working that is good 13:55 < amarchuk> and then we can agree on how to fix non-virtio things 13:56 < amarchuk> imo having a parameter might be reasonable since I do not yet have idea how to reliably detect it 13:59 < amarchuk> fromani, danken it is just two questions. So I suggest to get the patch merged since it makes what was not working - working without breaking anything else. And then when I have solid idea on how to handle non-virtio hosts I will create another patch for that. 14:01 <@danken> amarchuk: personally, I don't see a real need for isci/ide scratch pad. 14:01 <@danken> amarchuk: but please look at _makeName in vdsm/virt/vmdevices/storage.py for the modern-days way to assign a disk name. 14:01 < amarchuk> danken, will have a look 14:02 < amarchuk> danken, do you think there is way to reuse that code from hook without copy&paste? 14:03 <@danken> amarchuk: not really, I'm afraid. Hooks should better be independent of vdsm code 14:04 <@danken> amarchuk: also, please listen to fromani's suggestion regarding os.umask(). 14:04 < amarchuk> danken, than if we are just supporting virtio only it is easier. and just the documentation on the hook will need to be updated. 14:04 < amarchuk> danken, umask is not able to set +w it just can mask some permissions 14:04 <@danken> read `man 2 umask` to learn how to use it 14:05 <@danken> umask(0002) should make things writable to vdsm:kvm. isn't it? 14:06 < amarchuk> danken, did not work when I tried that. but ok. will study it more. maybe I was setting it in the wrong place 14:07 <@danken> amarchuk: should be set before creating the file 14:08 -!- bala (purple): has joined #vdsm 14:11 -!- ndarshan: has quit [Quit: Leaving] 14:11 < amarchuk> danken, ok will handle that. as for the device name code. it looks similar to storage.py providing that 1. we support virtio only so know that it is "vd" and 2. we may already have devices assigned so additionaly collect them and check that we are not creating a duplicate 14:14 -!- mpolednik: has quit [Ping timeout: 246 seconds] 14:29 -!- mpolednik (mpolednik): has joined #vdsm 14:31 -!- bala: has quit [Ping timeout: 252 seconds] 14:32 -!- sshnaidm: has quit [Ping timeout: 276 seconds] 14:33 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 14:39 -!- mpolednik: has quit [Ping timeout: 276 seconds] 14:41 -!- tjeyasin: has quit [Ping timeout: 246 seconds] 14:42 -!- timothy_: has quit [Ping timeout: 264 seconds] 14:42 -!- saggi (purple): has joined #vdsm 14:42 -!- _tim_07: has quit [Ping timeout: 272 seconds] 14:45 -!- bala (purple): has joined #vdsm 14:45 -!- mpolednik (mpolednik): has joined #vdsm 14:49 -!- ibarkan (purple): has joined #vdsm 14:51 -!- ykaplan: has quit [Ping timeout: 256 seconds] 14:54 -!- mpolednik: has quit [Ping timeout: 276 seconds] 15:00 -!- mpolednik (mpolednik): has joined #vdsm 15:06 -!- gshereme: has quit [Quit: Leaving] 15:11 -!- Madhu_: has quit [Remote host closed the connection] 15:12 -!- apahim_: has quit [Ping timeout: 246 seconds] 15:27 < alitke> nsoffer, So we may need to extend raw block volumes prior to live merge. 15:27 < alitke> Sadly this is a mailbox operation. Any ideas about how to integrate it with the synchronous part (beginning) of the live merge flow? 15:28 < alitke> We could require engine to manage the extend for us. 15:28 < alitke> In that case, I'll just add a new check in the merge verb and report an error when top is larger than base and base is a raw block volume. 15:28 -!- apahim_ (Amador Pahim): has joined #vdsm 15:29 < nsoffer> alitke, sounds reasonable, engine ask to do live merge, let it prepare the volume if needed 15:30 < alitke> ok... And also this isn't related to thin provisioning so we are not leaking that knowledge to engine. 15:30 < nsoffer> alitke, otherwise we are again making large vdsm flows which are hard to manage (what was failed, how to handle the failure) 15:30 < alitke> right. And as long as vdsm detects the situation and prevents it we should be good. 15:32 < alitke> cool. Now I just have to find gpadgett to talk about the engine side :) 15:33 < nsoffer> right, this is less cool for engine guys :-) 15:33 -!- ykaplan (Yeela Kaplan): has joined #vdsm 15:33 < nsoffer> alitke, I wonder if they know the size of the inner volume - do we report it? 15:34 < alitke> They track it in the DB. That is the other side of the problem/ 15:34 < alitke> When this scenario happens and the merge completes, they report the disk size according to the cached value for the base volume. 15:35 < alitke> THis ends up making it apppear as if the disk shrank to it's old size even though the storage is correct. 15:35 < nsoffer> alitke, if engine tracks the size of each snapshot, it can prepare for the live merge easily 15:35 < alitke> yes... Just need to confirm it with Greg. 15:36 < nsoffer> alitke, you can also talk with Daniel/Liron/Maor 15:48 < ibarkan> danken: phoracek: this patch https://gerrit.ovirt.org/#/c/42022/ will have a better error message when a functional test fails on a KernelConfig issue 15:53 -!- acanan: has quit [Quit: Leaving] 15:53 -!- acanan (Aharon Canan): has joined #vdsm 15:59 -!- gpadgett (Greg Padgett): has joined #vdsm 16:00 -!- dyasny (Dan Yasny): has joined #vdsm 16:00 -!- ibarkan: has quit [Ping timeout: 256 seconds] 16:03 -!- bazulay: has quit [Quit: Leaving.] 16:03 -!- bazulay (purple): has joined #vdsm 16:07 -!- ishaby: has quit [Ping timeout: 246 seconds] 17:09 -!- bala: has quit [Quit: Leaving.] 17:11 -!- sshnaidm: has quit [Ping timeout: 276 seconds] 17:15 < amarchuk> danken, I found why umask is not working. due to this https://github.com/qemu/qemu/blob/a3206972a9eab65ec8e8f9ae320ad628ba4b58f1/block/raw-posix.c#L492 qemu hardcode 0644 permissions and since umask is mode & ~umask we cannot really add any bits to it, just remove :-( 17:18 -!- acanan: has quit [Ping timeout: 248 seconds] 17:24 -!- ykaplan: has quit [Ping timeout: 252 seconds] 17:26 -!- _tim_07 (Tim): has joined #vdsm 17:34 -!- sbonazzo: has quit [Quit: Leaving.] 17:44 -!- fabiand: has quit [Ping timeout: 256 seconds] 17:46 -!- fabiand (Fabian Deutsch): has joined #vdsm 17:47 -!- pkliczew: has quit [Ping timeout: 265 seconds] 17:47 -!- _tim_07: has quit [Ping timeout: 272 seconds] 17:49 -!- nsoffer: has quit [Ping timeout: 264 seconds] 17:50 -!- ykaplan (Yeela Kaplan): has joined #vdsm 17:51 -!- gshereme (Greg Sheremeta): has joined #vdsm 17:51 -!- _tim_07 (Tim): has joined #vdsm 17:59 -!- bazulay: has quit [Ping timeout: 256 seconds] 18:09 -!- phoracek: has quit [Quit: WeeChat 1.1.1] 18:22 -!- jbrooks (Jason Brooks): has joined #vdsm 18:23 -!- fromani: has quit [Quit: Leaving] 18:29 -!- mpolednik: has quit [Ping timeout: 264 seconds] 18:35 -!- ykaplan: has quit [Remote host closed the connection] 18:50 -!- #vdsm dyasny: has quit [Ping timeout: 272 seconds] 18:54 -!- apahim_ is now known as apahim 18:56 -!- _tim_07: has quit [Ping timeout: 256 seconds] 19:26 -!- mpolednik (mpolednik): has joined #vdsm 19:31 -!- mpolednik: has quit [Ping timeout: 252 seconds] 20:41 -!- fabiand: has quit [Ping timeout: 256 seconds] 20:41 -!- fabiand (Fabian Deutsch): has joined #vdsm 21:09 -!- #vdsm ybronhei: has quit [Ping timeout: 265 seconds] 21:24 -!- dkuznets (Dima Kuznetsov): has joined #vdsm 21:24 -!- fabiand: has quit [Quit: Verlassend] 21:28 -!- mpolednik (mpolednik): has joined #vdsm 21:33 -!- mpolednik: has quit [Ping timeout: 264 seconds] 21:41 -!- dkuznets: has quit [Ping timeout: 248 seconds] 21:57 -!- dkuznets (Dima Kuznetsov): has joined #vdsm 22:36 -!- danken: has quit [Quit: Leaving.] 22:40 -!- dkuznets: has quit [Ping timeout: 250 seconds] 23:07 -!- nsoffer (Nir Soffer): has joined #vdsm 23:29 -!- mpolednik (mpolednik): has joined #vdsm 23:35 -!- mpolednik: has quit [Ping timeout: 246 seconds] 23:38 -!- mpolednik (mpolednik): has joined #vdsm 23:40 -!- mpolednik: has quit [Client Quit] 23:44 -!- nsoffer: has quit [Ping timeout: 246 seconds] 23:54 -!- mmirecki: has quit [Ping timeout: 265 seconds] --- Log closed do jun 25 00:00:00 2015