--- Log opened zo nov 24 00:00:59 2013 00:10 -!- moolit: has quit [Quit: Leaving.] 00:13 -!- celebdor: has quit [Ping timeout: 272 seconds] 00:29 -!- moolit (purple): has joined #vdsm 00:29 -!- nsoffer: has quit [Read error: Connection reset by peer] 00:46 -!- nsoffer (purple): has joined #vdsm 00:55 -!- moolit: has left #vdsm [] 03:06 -!- fabiand__ (Fabian Deutsch): has joined #vdsm 03:09 -!- fabiand_: has quit [Ping timeout: 246 seconds] 03:46 -!- nsoffer: has quit [Quit: Leaving.] 03:46 -!- nsoffer (purple): has joined #vdsm 03:50 -!- nsoffer: has quit [Client Quit] 04:21 -!- osier (osier): has joined #vdsm 05:33 -!- dougsland (Douglas): has joined #vdsm 05:33 -!- dougsland (Douglas): has joined #vdsm 06:45 -!- abaron: has quit [Ping timeout: 248 seconds] 06:58 -!- mohankumar (M. Mohan Kumar): has joined #vdsm 07:49 -!- abaron (Ayal Baron): has joined #vdsm 07:53 -!- shaohef: has quit [Quit: Leaving.] 07:53 -!- shaohef (shaohef): has joined #vdsm 07:55 -!- abaron: has quit [Ping timeout: 245 seconds] 08:40 -!- ewarszaw (Eduardo Warszawski): has joined #vdsm 08:46 -!- celebdor (celebdor): has joined #vdsm 08:52 -!- amuller (Assaf Muller): has joined #vdsm 09:09 -!- acanan (Aharon Canan): has joined #vdsm 09:14 -!- ybronhei: has quit [Quit: Leaving.] 09:34 -!- moolit (purple): has joined #vdsm 09:44 -!- moolit: has left #vdsm [] 09:49 -!- nsoffer (purple): has joined #vdsm 09:50 -!- saggi (purple): has joined #vdsm 09:52 -!- ybronhei (purple): has joined #vdsm 09:59 -!- moolit (purple): has joined #vdsm 10:08 -!- bazulay (purple): has joined #vdsm 10:08 -!- nsoffer: has quit [Read error: Connection reset by peer] 10:09 -!- ykaplan (Yeela Kaplan): has joined #vdsm 10:09 -!- nsoffer (purple): has joined #vdsm 10:09 -!- ybronhei: has quit [Read error: Connection reset by peer] 10:09 -!- ybronhei (purple): has joined #vdsm 10:10 -!- abaron (Ayal Baron): has joined #vdsm 10:10 < celebdor> amuller: ping 10:10 -!- ybronhei: has quit [Client Quit] 10:10 < amuller> celebdor: Heya 10:11 < amuller> What's up? 10:15 < celebdor> amuller: network status updated ;-) 10:15 < amuller> celebdor: Thanks 10:15 < celebdor> amuller: did you see what a ton of patches I got merged thanks to Dan the last three days? 10:16 < amuller> oh, wow 10:16 < amuller> I was meaning to sit down and review everything late today or tomorrow morning 10:16 < celebdor> unfortunately it got me so busy that I didn't have time for finishing to review the test 10:16 < celebdor> bowling test 10:17 < amuller> yeah I'm working on that now 10:18 < celebdor> I'd like a rephrase 10:18 -!- nsoffer: has quit [Read error: Connection reset by peer] 10:18 < celebdor> "The challenge that has been bestowed upon you will be a single test" 10:18 < celebdor> heh 10:18 < celebdor> :-) 10:19 < amuller> Done 10:19 < celebdor> :O It was a joke :P 10:19 < amuller> Undone. 10:19 < celebdor> I'm sure Dan wouldn't approve 10:19 < amuller> :) 10:19 < celebdor> but if it were me... 10:20 < amuller> I'm trying to come up with a phase 2 for the exam right now 10:20 < amuller> An additional test that requires more monkey patching 10:20 < celebdor> I wouldn't have included the runTests file 10:20 < celebdor> but I don't mind that much 10:21 < amuller> It's a possibility... force them to learn how to use the unittest library 10:35 -!- ybronhei (purple): has joined #vdsm 10:35 -!- nsoffer (purple): has joined #vdsm 10:36 -!- bazulay: has left #vdsm [] 10:37 -!- bazulay (purple): has joined #vdsm 11:14 -!- abaron: has quit [Remote host closed the connection] 12:31 -!- ykaplan: has quit [Ping timeout: 245 seconds] 13:03 -!- nsoffer: has quit [Quit: Leaving.] 13:35 -!- nsoffer (purple): has joined #vdsm 14:10 -!- ykaplan (Yeela Kaplan): has joined #vdsm 14:52 -!- ewarszaw: has quit [Quit: Konversation terminated!] 14:59 < celebdor> saggi: ping 15:00 < celebdor> to make sync IO with the asyncproc in vdsm.utils I had to do what's on the bottom of http://gerrit.ovirt.org/#/c/21430/4/lib/vdsm/ipwrapper.py 15:00 -!- nsoffer: has quit [Read error: Connection reset by peer] 15:01 < celebdor> using epoll to check the on the proc stdout 15:01 < celebdor> Dan was wondering if asyncproc could be tweaked to allow sync operation in one of the streams by itself (witohout using an extra method like I did) 15:02 * celebdor is apuimedo 15:02 -!- nsoffer (purple): has joined #vdsm 15:03 < celebdor> nsoffer: did you see that now it only happens once per every 2 seconds? 15:03 < celebdor> (it will be completely gone soon) 15:05 -!- ykaplan: has quit [Read error: Operation timed out] 15:08 -!- saggi: has quit [Ping timeout: 246 seconds] 15:14 < nsoffer> celebdor: what? 15:30 -!- saggi (purple): has joined #vdsm 15:30 < saggi> celebdor: ? 15:32 < celebdor> saggi: http://gerrit.ovirt.org/#/c/21430/4/lib/vdsm/ipwrapper.py 15:32 < celebdor> here _stdoutLines is something that I did to have synced IO over stdout of an utils.AsyncProc 15:33 < celebdor> danken was wondering if it were possible to tweak asyncproc so it is possible to get sync io from a single stream 15:33 < celebdor> so that my code would not be necessary 15:35 -!- saggi: has quit [Quit: Leaving.] 15:37 -!- saggi (purple): has joined #vdsm 15:40 -!- nsoffer: has quit [Read error: Connection reset by peer] 15:41 -!- nsoffer (purple): has joined #vdsm 15:43 -!- saggi: has quit [Quit: Leaving.] 15:43 -!- nsoffer: has quit [Read error: Connection reset by peer] 15:43 -!- nsoffer (purple): has joined #vdsm 15:45 < celebdor> nsoffer: that's what I was asking saggi, for being able asyncproc with sync io 15:45 < celebdor> danken and I couldn't figure it out 15:52 -!- apuimedo (antoni): has joined #vdsm 15:52 -!- nsoffer: has quit [Read error: Connection reset by peer] 15:54 -!- nsoffer (purple): has joined #vdsm 16:04 < amuller> ybronhei: http://gerrit.ovirt.org/#/c/21590/ 16:04 < amuller> ybronhei: patch that adds touch 16:06 < ybronhei> amuller: thanks :) 16:25 -!- #vdsm,#openshift osier: has quit [Ping timeout: 264 seconds] 16:42 < nsoffer> celebdor: will http://pastebin.com/q5bcdmBZ work for the ip monitor? 16:45 < celebdor> nsoffer: nope 16:46 < celebdor> when you pass sync=false to execCmd 16:46 < celebdor> it makes an AsyncProc 16:46 < celebdor> for which the io operations are async too 16:46 < celebdor> so when you read and the buffer is empty it returns '' 16:47 < celebdor> making the iteration through lines stop 16:47 < celebdor> (that's why I use the system poll) 16:47 < nsoffer> my test prove that it does not 16:47 < nsoffer> try it 16:47 < celebdor> I tried it on my machine last week 16:48 < celebdor> for ip monitor it didn't work 16:49 -!- nsoffer: has quit [Read error: Connection reset by peer] 16:51 -!- nsoffer (purple): has joined #vdsm 16:51 < nsoffer> celebdor: check http://pastebin.com/m7btQRRM 16:52 < nsoffer> It could be problem with output buffering in ip 16:52 < celebdor> nsoffer: try putting in monitor.sh a sleep before the iteration of 7" 16:54 < celebdor> iirc it had relation to finding the buffer empty from the beggining for several tries 16:58 < nsoffer> celebdor: yea, it does work 16:58 < nsoffer> I mean, doe not :-) 17:00 < celebdor> :( It's a pity 17:00 -!- nsoffer: has quit [Read error: Connection reset by peer] 17:01 -!- nsoffer (purple): has joined #vdsm 17:06 < celebdor> nsoffer: abaron's comment is about the extra coupling vm -> hsm introduced by passing the callback that touches the internal vm.py data, right? 17:07 < nsoffer> I don't see any touching of internal vm.py data, but thats probably what he meant 17:07 < celebdor> it's a bit of a nasty side effect, tbh 17:07 < nsoffer> It is a callback - what the caller does with it is his business 17:07 < nsoffer> I don't see any coupling 17:08 < celebdor> well, that was abaron's concern 17:08 < celebdor> or that's how I interpreted it 17:08 < nsoffer> If you create an object that is not ready for use, you want to know when it is ready 17:08 -!- abaron (Ayal Baron): has joined #vdsm 17:09 < celebdor> nsoffer: I was thinking about futures 17:09 < celebdor> In some Java project I was using them for a similar situation 17:09 < celebdor> but I think they are only in Py3k 17:10 < nsoffer> lets say we have futures, how can they help with this? 17:10 < celebdor> the end result would be the same 17:11 < celebdor> just stdlibish 17:12 < celebdor> nsoffer: what about in the ready property you just ask irs if it is ready itself 17:12 < celebdor> that wouldn't need the callback, would it? 17:12 < nsoffer> I had one in patch 1 - but it is hidden by the Dispatcher wrapping HSM 17:13 < nsoffer> and then you have to ask it on every call if it is ready - good if it could switch from not-ready to ready state while running 17:13 < celebdor> HSM could have a initialized property 17:13 < celebdor> that clientif asks in ready 17:13 < nsoffer> but it does not - it go once from not-ready to read 17:14 < celebdor> well, there's no reason the property could not set the self._irsReady to True (originally having it None) 17:14 < celebdor> and if it is that it doesn't need to do anything else 17:15 < celebdor> @property 17:15 < celebdor> (nevermind the last line 17:15 < celebdor> ) 17:16 < nsoffer> but then hsm.ready should be public 17:16 < nsoffer> or you cannot access it though the dispatcher 17:16 < nsoffer> or we change the dispatcher - seem to much work for no value 17:17 * celebdor hasn't looked at the dispatcher :P 17:17 < celebdor> I don't know how much work would that be 17:17 < celebdor> but it seems like a dispatcher should have some notion of the status of what it dispatches 17:17 < nsoffer> The dispatcher replace @public method with protected methods 17:18 < nsoffer> and hiding anything which is not public 17:18 -!- osier (osier): has joined #vdsm 17:19 < celebdor> nsoffer: maybe it's a silly question, but... Why all the hiding? 17:19 < nsoffer> I don't know :-) 17:19 < celebdor> sounds very unpythonic and very controlfreakish to me 17:19 < celebdor> there must be a very good and undocumented reason 17:21 < celebdor> maybe abaron could shed some light as to why it is done 17:21 -!- nsoffer: has quit [Read error: Connection reset by peer] 17:21 -!- nsoffer (purple): has joined #vdsm 17:21 < abaron> celebdor: ??? 17:22 < celebdor> abaron: I was curious what is the motivation behind the shadowing and wrapping done by the storage dispatcher 17:22 < nsoffer> celebdor: so in this patch, that callback is the only thing that does not look right? 17:22 < abaron> ah, ancient history. 17:22 < abaron> it actually started even before I got into it. 17:22 < celebdor> abaron: yeah, I was hoping you had a ready answer before I exercised the old logs 17:22 < abaron> celebdor: I doubt you'd be able to find anything there 17:23 < celebdor> nsoffer: to me yes, the callback that touches the internal clientif thing is the only thing that makes me wonder if we could make it in a simpler way 17:24 < celebdor> abaron: not even before "tabula rasa"? 17:24 < nsoffer> celebdor: why do you care what that callback does? 17:25 < nsoffer> I mean, every object method touch that object internal state 17:25 < celebdor> nsoffer: for the selfish reason that if I were to find out the flow of the process it would take more work 17:25 < celebdor> I generally dislike travelling instance methods 17:25 < nsoffer> celebdor: thats how async flows work 17:25 < nsoffer> that life 17:25 < celebdor> :P 17:26 < nsoffer> of course we could block in the caller on an event 17:26 < celebdor> I would prefer the dispatcher getting axed and the ready attribute to be on HSM and the clientif ready property checking, as I said before 17:26 < celebdor> :P 17:26 < nsoffer> but the whole point of this is start serving now, no blocking 17:27 < celebdor> what I propose doesn't block 17:27 < celebdor> but it is more work and we don't know if we can get rid of the Dispatcher 17:27 < celebdor> until we can do that 17:27 < celebdor> I don't have anything against your patch 17:28 < celebdor> (except the typo I put in the comments) 17:28 < celebdor> :P 17:28 -!- mohankumar: has quit [Ping timeout: 246 seconds] 17:28 < nsoffer> Well what you describe it he first version, before I checked what the dispatcher does 17:29 < nsoffer> It seems the simplest solution that can work 17:29 < nsoffer> but it did not :-) 17:30 < celebdor> understood 17:30 -!- nsoffer: has quit [Read error: Connection reset by peer] 17:31 -!- nsoffer (purple): has joined #vdsm 17:33 < celebdor> nsoffer: office connection disconnecting you over and over? 17:33 -!- amuller: has quit [Quit: Konversation terminated!] 17:33 < nsoffer> home office :-) 17:34 < nsoffer> It is only irc connection - I don't see problem in any other app 17:34 < nsoffer> maybe something with the vpn? 17:35 < celebdor> nsoffer: could be 17:35 < celebdor> which vpn? TLV's? 17:35 < nsoffer> tlv 17:35 < celebdor> yeah... Not a very good one, from my experience there 17:35 < nsoffer> or it is my irc client (pidgin) 17:36 < celebdor> AMS works so much better 17:36 < celebdor> nsoffer: weechat FTW 17:45 -!- nsoffer1 (nsoffer): has joined #vdsm 17:48 -!- bazulay: has quit [Quit: Leaving.] 17:49 -!- ykaplan (Yeela Kaplan): has joined #vdsm 17:51 -!- Tankado (realname): has joined #vdsm 17:57 -!- nsoffer: has quit [Read error: Connection reset by peer] 17:58 -!- nsoffer (purple): has joined #vdsm 18:03 -!- nsoffer1: has quit [Quit: WeeChat 0.4.2] 18:03 -!- nsoffer: has quit [Read error: Connection reset by peer] 18:03 -!- apuimedo: has quit [Read error: Connection reset by peer] 18:03 -!- nsoffer (nsoffer): has joined #vdsm 18:04 -!- nsoffer1 (purple): has joined #vdsm 18:04 -!- ykaplan: has quit [Ping timeout: 272 seconds] 18:06 -!- nsoffer: has quit [Client Quit] 18:07 -!- nsoffer (Nir Soffer): has joined #vdsm 18:09 -!- Tankado: has left #vdsm ["Leaving"] 18:14 -!- nsoffer: has left #vdsm ["Leaving"] 18:14 -!- nsoffer (Nir Soffer): has joined #vdsm 18:14 -!- acanan: has quit [Ping timeout: 264 seconds] 18:16 -!- nsoffer1: has quit [Quit: Leaving.] 18:18 -!- nsoffer: has quit [Read error: Connection reset by peer] 18:20 -!- ybronhei: has quit [Quit: Leaving.] 19:20 -!- moolit: has quit [Ping timeout: 252 seconds] 19:20 -!- ybronhei (purple): has joined #vdsm 19:21 -!- ybronhei: has quit [Client Quit] 19:51 -!- nsoffer (purple): has joined #vdsm 19:51 -!- nsoffer_ (Nir Soffer): has joined #vdsm 19:59 -!- nsoffer_: has quit [Quit: Leaving] 20:07 -!- nsoffer_ (Nir Soffer): has joined #vdsm 20:12 -!- nsoffer_: has quit [Quit: Ex-Chat] 20:31 -!- nsoffer: has quit [Quit: Leaving.] 20:34 -!- nsoffer (purple): has joined #vdsm 20:39 -!- nsoffer: has quit [Quit: Leaving.] 20:43 -!- nsoffer (Nir Soffer): has joined #vdsm 20:45 -!- moolit (purple): has joined #vdsm 21:09 -!- ybronhei (purple): has joined #vdsm 21:10 -!- ybronhei: has quit [Client Quit] 21:15 -!- nsoffer: has left #vdsm [] 21:16 -!- nsoffer1 (purple): has joined #vdsm 21:16 -!- nsoffer1: has quit [Client Quit] 21:17 -!- nsoffer (Nir Soffer): has joined #vdsm 21:17 -!- abaron: has quit [Ping timeout: 264 seconds] 21:27 -!- #vdsm,#openshift osier: has quit [Ping timeout: 252 seconds] 21:52 -!- moolit: has quit [Quit: Leaving.] 22:41 -!- nsoffer: has quit [Read error: Connection reset by peer] 22:45 -!- nsoffer (Nir Soffer): has joined #vdsm 23:12 -!- nsoffer: has quit [Ping timeout: 246 seconds] --- Log closed ma nov 25 00:00:01 2013