--- Log opened do jun 25 00:00:00 2015 00:30 -!- #vdsm dougsland: has quit [Ping timeout: 250 seconds] 00:41 -!- adahms (Andrew Dahms): has joined #vdsm 00:42 -!- dougsland (Douglas): has joined #vdsm 01:28 -!- saggi: has quit [Ping timeout: 246 seconds] 01:38 -!- saggi (purple): has joined #vdsm 01:56 -!- #vdsm dougsland: has quit [Ping timeout: 252 seconds] 02:04 -!- dougsland (Douglas): has joined #vdsm 02:43 -!- bala (purple): has joined #vdsm 03:04 -!- #vdsm dougsland: has quit [Ping timeout: 264 seconds] 03:16 -!- bala: has quit [Read error: Connection reset by peer] 03:30 -!- bala (purple): has joined #vdsm 03:36 -!- gpadgett: has quit [Quit: Leaving] 04:02 -!- bala: has quit [Read error: No route to host] 05:41 -!- vered: has quit [Ping timeout: 255 seconds] 05:51 -!- dyasny (Dan Yasny): has joined #vdsm 06:06 -!- shubhendu (Shubhendu Ram Tripathi): has joined #vdsm 06:55 -!- vered (Vered Volansky): has joined #vdsm 07:11 -!- ndarshan (Darshan n): has joined #vdsm 07:13 -!- #vdsm dyasny: has quit [Ping timeout: 244 seconds] 07:27 -!- Madhu_ (Madhu_): has joined #vdsm 07:34 -!- fabiand (Fabian Deutsch): has joined #vdsm 07:37 -!- mmirecki (Marcin Mirecki): has joined #vdsm 08:08 -!- nsoffer (Nir Soffer): has joined #vdsm 08:11 -!- sbonazzo (purple): has joined #vdsm 08:16 -!- ishaby (Idan Shaby): has joined #vdsm 08:28 -!- dkuznets (Dima Kuznetsov): has joined #vdsm 08:31 -!- timothy_ (Timothy Asir): has joined #vdsm 08:31 -!- tjeyasin (Timothy Asir): has joined #vdsm 08:33 -!- fromani (Francesco Romani): has joined #vdsm 08:34 -!- _tim_07 (Tim): has joined #vdsm 08:38 -!- pkliczew (Piotr Kliczewski): has joined #vdsm 08:44 -!- shubhendu: has quit [Quit: Leaving] 08:46 -!- mpolednik (mpolednik): has joined #vdsm 08:50 -!- mpolednik: has quit [Client Quit] 08:56 -!- mpolednik (mpolednik): has joined #vdsm 09:00 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 09:06 -!- shubhendu (Shubhendu Ram Tripathi): has joined #vdsm 09:09 -!- sshnaidm: has quit [Ping timeout: 276 seconds] 09:13 -!- ybronhei (purple): has joined #vdsm 09:20 -!- adahms: has quit [Quit: Leaving] 09:22 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 09:24 < mpolednik> fromani: https://gerrit.ovirt.org/#/c/42757/ 09:24 < mpolednik> evil test removal, but since we rely on libvirt now it doesn't make sense to use that kind of test 09:33 < fromani> mpolednik: looking 09:43 -!- phoracek (phoracek): has joined #vdsm 09:50 -!- mode/#vdsm: by ChanServ 09:50 -!- danken (purple): has joined #vdsm 09:57 -!- acanan (Aharon Canan): has joined #vdsm 09:57 -!- acanan is now known as acanan_WFH 09:57 -!- acanan_WFH is now known as acanan 09:59 -!- mpolednik1 (mpolednik): has joined #vdsm 10:00 -!- mpolednik: has quit [Ping timeout: 255 seconds] 10:06 < fromani> mpolednik1: yay, looks ok! 10:06 -!- mpolednik1: has quit [Ping timeout: 246 seconds] 10:07 -!- sshnaidm: has quit [Ping timeout: 276 seconds] 10:12 -!- phoracek: has quit [Ping timeout: 246 seconds] 10:20 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 10:20 < fabiand> fromani, hey 10:20 < fromani> fabiand: hi, morning! 10:20 < fabiand> good morning fromani :) 10:22 < fromani> fabiand: :) 10:23 < fabiand> fromani, I thought I rememberd that we spoke about bonds in the past 10:23 < fabiand> but I think it was vlad .. 10:24 < fromani> fabiand: maybe in a distant past, can't recall anything... 10:26 -!- ykaplan (Yeela Kaplan): has joined #vdsm 10:27 -!- phoracek (phoracek): has joined #vdsm 10:42 -!- ibarkan (purple): has joined #vdsm 10:44 < ibarkan> danken: ybronhei: make check is green and happy on master... 10:47 < ibarkan> fabiand: about https://bugzilla.redhat.com/show_bug.cgi?id=1225972 10:48 < ibarkan> fabiand: I don't have the bw atm. 3.5.4 is on my back. I'll get to it in a few days, I hope. 10:48 < fabiand> ibarkan, hey - green is good! - yes (about the bug)? 10:48 < ibarkan> fabiand: bw=bandwidth 10:48 < fabiand> okay - shall I set needinfo on danken so he can take a look at it as well? 10:48 < fabiand> I am just not sure if this is something which can lead to problems 10:48 < fabiand> I've not seen the message before 10:49 < ibarkan> fabiand: me neither. it needs to be looked at. but it is in low priority right now for me. 10:49 < fabiand> okay 10:49 -!- phoracek: has quit [Ping timeout: 246 seconds] 10:50 < fabiand> ibarkan, no worries 10:50 < fabiand> ibarkan, but thanks for heads up 11:00 -!- Madhu_: has quit [Quit: Madhu_] 11:03 -!- phoracek (phoracek): has joined #vdsm 11:04 -!- bazulay (purple): has joined #vdsm 11:07 < amarchuk> danken fromani , ping. Let's see what is left on https://gerrit.ovirt.org/#/c/42573 1. umask is inevstigated and we cannot use it due to qemu-img permissions hardcode so chmod is ok. 2. device enumeration code is similar to vdsm/virt/vmdevices/storage.py since we are not reusing it in hooks copy past is ok. 3. virtio I would do as danken said and just do not support non-virtio unless somebody will need it in future. we can update wiki. So n 11:07 < amarchuk> o other pending things left with the patch 11:08 < fromani> amarchuk: yep looks like that's all for version3 11:10 < amarchuk> fromani, yeah and thins virtio is already in the master anyway by merging the fix I believe we will enable support for most installations that is good enough right now. As I said we are going to be a major user for that hook so if there are any more problems we will uncover them for sure 11:11 < nsoffer> ibarkan, ping 11:11 < ibarkan> nsoffer: hi 11:11 < nsoffer> ibarkan, I get two consistent test failures on master 11:11 < ibarkan> nsoffer: which? 11:11 < nsoffer> testIterNetworkHierarchy, testTextualRepr 11:12 < nsoffer> ibarkan, I will paste the exception soon (still running) 11:12 < nsoffer> (check-all is slow) 11:13 < nsoffer> ibarkan, both fail because: IOError:: No such file or directory: '/usr/local/var/lib/vdsm/bonding-defaults.json' 11:13 < nsoffer> ibarkan, we fix this last week or so - didn't we? 11:13 < ibarkan> nsoffer: oh that 11:14 < ibarkan> nsoffer: we did not afaik 11:14 < ibarkan> vdsm restart should create it I think 11:14 <@danken> nsoffer: we only removed the file (then it failed due to bad permissions of an existing file. now it does not exist) 11:15 < ibarkan> danken: this is probably what ybronhei was talking about 11:15 < nsoffer> danken, so why the test is assuming that we have such file? 11:15 < nsoffer> danken, did you have to time to look at the ceph secret management patches? 11:16 < nsoffer> danken, in the call you said you are going to take them, but the line was very bad, maybe there was some misunderstanding :-) 11:17 <@danken> nsoffer: I am going to take them, they are on the top of my list for this morning. 11:17 < nsoffer> danken, great, ping me if you have a question 11:17 <@danken> ibarkan: btw, dkuznets reports a regression in master: http://fpaste.org/236458/43522279/ 11:18 <@danken> ibarkan please take a look on that, too. 11:24 < nsoffer> danken, and the first sdm refactorings - waiting for Federico? 11:31 < pkliczew> danken, hi 11:32 <@danken> ibarkan: dkuznets provided more logs: http://fpaste.org/236465/52239221/ 11:33 <@danken> pkliczew: hi! 11:33 <@danken> nsoffer: yes, but not forever. 11:33 < pkliczew> danken, I am enabled the vdsm_master_unit-tests_created job again 11:33 < ibarkan> danken: we already saw this 11:33 < pkliczew> danken, after the investigation it seems that it is related to CI env 11:34 < pkliczew> danken, now waiting on job get stuck to investigate 11:34 < pkliczew> danken, will let you know once we know what is the issue 11:35 <@danken> pkliczew: thanks. have you tried setting NOSE_EXCLUDE on the CI job, so that "your" tests are skipped? 11:36 < pkliczew> danken, it is running now, we need to reproduce the issue 11:36 <@danken> pkliczew: when the tests hangs up again, I'd like to skip them without disabling everything. 11:36 < pkliczew> danken, sure 11:36 <@danken> pkliczew: yeah, I just like you to be ready with the NOSE_EXCLUDE at your pocket. 11:36 < pkliczew> danken, sure 11:41 -!- Madhu_ (Madhu_): has joined #vdsm 11:43 -!- tjeyasin: has quit [Quit: Leaving] 11:43 -!- timothy_: has quit [Quit: Leaving] 11:48 -!- ykaplan: has quit [Ping timeout: 264 seconds] 11:57 <@danken> nsoffer: can you share more of the traceback of the bonding.json failure? 11:57 < nsoffer> danken, sure, let me paste it 11:58 < nsoffer> danken, ibarkan http://pastebin.com/BEVHUqXz 11:59 <@danken> nsoffer: thanks. ibarkan^^^ 12:12 < ibarkan> danken: nsoffer: this test should have the @MonkeyPatch(netinfo, ... code 12:17 <@danken> ibarkan we should make this monkey-patching on the module (or runner) level, I presume 12:18 < ibarkan> danken: yea 12:22 -!- nsoffer: has quit [Ping timeout: 252 seconds] 12:26 -!- ykaplan (Yeela Kaplan): has joined #vdsm 12:28 < ibarkan> danken: we have MonkeyClass for that 12:37 <@danken> ibarkan, which does this in the class level. 12:37 < ibarkan> danken: Test class level. so this is what we want 12:37 <@danken> ibarkan but I think that NOTHING should look for /va/rlib/bonding.json in ANY test. so the monkeypatching can happen much much higher 12:38 <@danken> ibarkan: but yes, moving this from a single test to the class is (a bit of) an improvement. 12:39 < ibarkan> danken: might be good enough for now. those are unit tests so it is not _that_ likely that it will happen elsewhere (and not impossible also) 12:55 -!- dougsland (Douglas): has joined #vdsm 13:02 < fromani> danken: in 3.6 we want to migrate using JSON-RPC only, right? I mean, no fallback to XML-RPC ever, correct? 13:13 -!- bala1 (purple): has joined #vdsm 13:21 -!- ibarkan2: has quit [Quit: AndroIRC - Android IRC Client ( http://www.androirc.com )] 13:22 <@danken> fromani: I don't think we can throw away xml rpc yet. 13:22 <@danken> pkliczew: ? 13:22 < pkliczew> danken, fromani as far as I know xmlrpc is fully supported in 3.6 13:23 < fromani> danken: I'm asking because I'm playing with migrations today and I find some inconsistencies in the xmlrpc vs jsonrpc path 13:23 < fromani> danken: pkliczew: I'll send a few patches once I sort them out so we can discuss ^^^ 13:23 < pkliczew> fromani, sure 13:24 -!- nsoffer (Nir Soffer): has joined #vdsm 13:29 < nsoffer> danken, thanks for the review, would you look at all the series? 13:31 <@danken> nsoffer: yeah, I looked up the smaller patches on top, they seem fine to me (thought I dislike slowing shutdown just to remove the secrets from libvirt) 13:31 <@danken> nsoffer: pkill libvirt would do the same ;-) 13:32 < nsoffer> danken, practically it is very fast 13:32 < nsoffer> danken, yea, but killing libvirt is really bad idea when you run vms 13:32 < nsoffer> danken, we can remove the shutdown cleanup, because we have startup cleanup as well 13:33 <@danken> nsoffer: I know... but I understand why you'd like to try to have a clean slate 13:34 <@danken> nsoffer: btw, what happens if Engine does not notice that Vdsm was down? 13:34 < nsoffer> danken, well you cannot start vms 13:34 < nsoffer> danken, but you will not able to do this anyway, since vdsm is not connected to storage 13:34 < nsoffer> danken, so as soon as engine connect vdsm back to storage, it will have the secrets again 13:36 <@danken> nsoffer: ok. 13:37 <@danken> nsoffer: and why's base64? do we need yet another level of marshaling? 13:38 -!- bala1: has quit [Read error: Connection reset by peer] 13:39 < nsoffer> danken, we get the secret encoded, why decode it in engine? 13:39 < nsoffer> danken, it is easier when we pass the same data all the way to libvirt 13:40 < nsoffer> danken, the only place where we must decode it is libvirt api 13:40 < nsoffer> so I don't see why we should deal with binary data when we have nicely ascii encoded string 13:50 <@danken> nsoffer: ok, if Engine's source is already armoured, it's convincing. 13:52 -!- sshnaidm: has quit [Ping timeout: 275 seconds] 13:55 -!- apuimedo (Antoni Segura Puimedon): has joined #vdsm 13:57 -!- ndarshan: has quit [Quit: Leaving] 13:59 < ibarkan> danken: nsoffer: do you know the --follow flag for nose ? 13:59 < nsoffer> ibarkan, no 14:00 < ibarkan> nsoffer: it's cool- it follows the suite run and records the failing tests 14:00 < ibarkan> nsoffer: the next run with the flag will tun only the failed tests 14:00 < ibarkan> run* 14:01 < nsoffer> ibarkan, do we need to enable it in testlib, or you can simply install it locally and enable it in the command line? 14:02 < ibarkan> nsoffer: nothing to install. it's a built-in plugin 14:04 -!- apahim: has quit [Ping timeout: 256 seconds] 14:11 -!- Madhu_: has quit [Quit: Madhu_] 14:20 -!- apahim (Amador Pahim): has joined #vdsm 14:37 < fromani> danken: pkliczew: I realized I worded my previous question very poorly. Let me retry: in 3.6, for VDSM<=>VDSM communication, most notably for migration, we'll use json-rpc only with no fallback option to xml-rpc, right? 14:37 < pkliczew> danken, yes 14:38 -!- fabiand: has quit [Ping timeout: 256 seconds] 14:47 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 14:50 <@danken> fromani: ykaplan's code (introducing jsonrpc) *did* add a fallback. 14:50 -!- fabiand (Fabian Deutsch): has joined #vdsm 14:50 <@danken> we can discuss if it makes sense, thought. 14:51 < fromani> danken: right. I was stuck thinking about an explicit fallback, e.g. checkbox button like we have in host config. in Engine. 14:52 <@danken> fromani: .me do not understand you. migrate() verb was not modified. it tries json, then falls to xml. 14:53 <@danken> fromani: hmmm, as it stands, 3.6 still supports clusterLevel 3.4, which did not have jsonrpc yet. 14:53 <@danken> fromani: so we need to keep this fallback. 14:53 < fromani> danken: sure sure, my fault. I had too narrow vision. Still, there can be some glitches worth investigating. I'm on it 14:56 -!- sshnaidm: has quit [Ping timeout: 276 seconds] 14:56 -!- phoracek: has quit [Quit: WeeChat 1.1.1] 14:56 -!- firemanxbr (Marcelo Barbosa): has joined #vdsm 15:00 < fromani> danken: here's the first link of my chain of thoughts: https://gerrit.ovirt.org/gitweb?p=vdsm.git;a=blob;f=lib/vdsm/jsonrpcvdscli.py;h=078b455a6cb2bda3487c6e03bf01004d58ae994f;hb=master#l68 . If I'm not mistaken, this means that this code will raise JsonRpcError on error, instead of returning an error response. 15:00 < fromani> danken: if it is so, then we must either fix this code to behave like xmlrpc vdscli, or let migration.py handle this. 15:01 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 15:06 -!- gpadgett (Greg Padgett): has joined #vdsm 15:09 <@danken> fromani: I do not follow. I thought the migrate.py is handling this in ykaplan's 47b611e9 15:11 < fromani> danken: different issue :) no worries, I'm on it 15:12 < ykaplan> fromani, do you need me to explain? 15:15 < fromani> ykaplan: not urgent :) but since you're here, a quick question. What happens if https://gerrit.ovirt.org/gitweb?p=vdsm.git;a=blob;f=lib/vdsm/jsonrpcvdscli.py;h=078b455a6cb2bda3487c6e03bf01004d58ae994f;hb=master#l64 fails? 15:15 < fromani> I mean, caller will get JsonRpcError, right? 15:15 < fromani> ykaplan: ^^^^ 15:16 < ykaplan> it depends what it fails on 15:20 < ykaplan> the cases in which we failover to xmlrpc are 2: 15:21 < ykaplan> 1. when destination communicates only by xmlrpc (we will get JsonRpcNoResponseError) 15:21 < ykaplan> 2. when source has no jsonrpc bindings (JsonRpcBindingsError) 15:21 < ykaplan> fromani, ^^^ 15:22 < fromani> ykaplan: ok with this 15:22 -!- sshnaidm: has quit [Ping timeout: 275 seconds] 15:23 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 15:25 -!- _tim_07: has quit [Ping timeout: 255 seconds] 15:26 < ykaplan> fromani, so you will get the JsonRpcError only in case something fails when you're already communicating with Json, and then we fail the migration and not fallover to xmlrpc... 15:26 < ykaplan> is that what you were looking for? 15:26 < fromani> ykaplan: yep 15:26 < ykaplan> is it not good enough? 15:26 < alitke> danken, Anything further I can do to convince you to take the first for StorageDomainManifest patches? 15:27 < ykaplan> fromani, cool! do you need me to fix something? 15:27 < fromani> ykaplan: it *is* cool enough, it is just not what migration.py expects. So the transport in use (JSON-RPC vs XML-RPC) is not really transparent here. 15:28 < fromani> s/cool/good 15:28 -!- apuimedo: has quit [Ping timeout: 264 seconds] 15:37 -!- sshnaidm: has quit [Ping timeout: 276 seconds] 15:38 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 15:58 -!- dyasny (Dan Yasny): has joined #vdsm 16:13 -!- ykaplan: has quit [Remote host closed the connection] 16:58 -!- #vdsm ybronhei: has quit [Ping timeout: 250 seconds] 17:05 -!- dyasny_ (Dan Yasny): has joined #vdsm 17:08 -!- pkliczew: has quit [Quit: Leaving] 17:08 -!- #vdsm dyasny: has quit [Ping timeout: 256 seconds] 17:13 -!- ishaby: has quit [Ping timeout: 265 seconds] 17:18 -!- shubhendu: has quit [Ping timeout: 244 seconds] 17:23 < fromani> headsup: when we'll upgrade to pep8 1.6.x, we'll have a lot of fun... 17:33 -!- dkuznets: has quit [Ping timeout: 248 seconds] 17:33 -!- acanan: has quit [Remote host closed the connection] 17:43 -!- fromani: has quit [Quit: Leaving] 17:56 -!- vered: has quit [Ping timeout: 264 seconds] 17:57 -!- sshnaidm: has quit [Ping timeout: 276 seconds] 17:58 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 17:59 -!- sbonazzo: has quit [Quit: Leaving.] 18:08 -!- ybronhei (purple): has joined #vdsm 18:17 -!- #vdsm ybronhei: has quit [Ping timeout: 272 seconds] 18:33 -!- sshnaidm: has quit [Ping timeout: 277 seconds] 18:33 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 19:16 -!- jbrooks: has quit [Read error: Connection reset by peer] 19:21 -!- sshnaidm: has quit [Ping timeout: 276 seconds] 19:22 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 19:28 -!- jbrooks (Jason Brooks): has joined #vdsm 19:31 -!- _tim_07 (Tim): has joined #vdsm 19:33 -!- jbrooks_ (Jason Brooks): has joined #vdsm 19:35 -!- sshnaidm: has quit [Ping timeout: 276 seconds] 19:36 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 19:36 -!- jbrooks: has quit [Ping timeout: 246 seconds] 19:44 -!- fabiand: has quit [Quit: Verlassend] 19:47 -!- jbrooks_ is now known as jbrooks 20:27 -!- dkuznets (Dima Kuznetsov): has joined #vdsm 20:32 -!- sshnaidm: has quit [Ping timeout: 276 seconds] 20:34 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 20:45 -!- bazulay: has quit [Ping timeout: 264 seconds] 21:00 -!- Madhu_ (Madhu_): has joined #vdsm 21:05 -!- sshnaidm: has quit [Ping timeout: 276 seconds] 21:06 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 21:21 -!- sshnaidm: has quit [Ping timeout: 277 seconds] 21:22 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 21:31 -!- ibarkan_ (ibarkan): has joined #vdsm 21:39 -!- ibarkan: has left #vdsm [] 21:42 -!- danken: has quit [Quit: Leaving.] 21:47 -!- Madhu_: has quit [Quit: Madhu_] 21:54 -!- sshnaidm: has quit [Ping timeout: 276 seconds] 21:55 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 22:03 -!- ibarkan_: has left #vdsm ["WeeChat 1.0.1"] 22:04 -!- ibarkan_ (ibarkan): has joined #vdsm 22:20 -!- sshnaidm: has quit [Ping timeout: 276 seconds] 22:20 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 22:22 -!- ibarkan_: has quit [Ping timeout: 264 seconds] 22:28 -!- firemanxbr: has quit [Quit: Leaving] 22:59 -!- sshnaidm: has quit [Ping timeout: 276 seconds] 22:59 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 23:08 -!- sshnaidm: has quit [Ping timeout: 276 seconds] 23:09 -!- _tim_07: has quit [Ping timeout: 272 seconds] 23:09 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 23:16 -!- sshnaidm: has quit [Ping timeout: 276 seconds] 23:16 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 23:19 -!- dkuznets: has quit [Ping timeout: 272 seconds] 23:26 -!- sshnaidm: has quit [Ping timeout: 277 seconds] 23:27 -!- sshnaidm (Sergey (Sagi) Shnaidman): has joined #vdsm 23:53 -!- mpolednik1 (mpolednik): has joined #vdsm --- Log closed vr jun 26 00:00:01 2015