Friday, 2013-12-13

*** kgriffs_afk is now known as kgriffs00:06
*** kgriffs is now known as kgriffs_afk00:19
*** openstackgerrit has quit IRC00:24
*** openstackgerrit has joined #openstack-marconi00:24
*** ekarlso has quit IRC00:25
*** ekarlso has joined #openstack-marconi00:25
*** reed has quit IRC00:28
*** haomaiwang has joined #openstack-marconi01:09
*** kgriffs_afk is now known as kgriffs01:11
*** haomaiwang has quit IRC01:14
*** kgriffs is now known as kgriffs_afk01:21
*** kgriffs_afk is now known as kgriffs02:12
*** kgriffs is now known as kgriffs_afk02:22
*** reed has joined #openstack-marconi02:37
*** haomaiwang has joined #openstack-marconi03:11
*** kgriffs_afk is now known as kgriffs03:13
*** haomaiwang has quit IRC03:16
*** flwang has joined #openstack-marconi03:20
*** kgriffs is now known as kgriffs_afk03:22
*** cpallares has quit IRC03:29
*** cpallares has joined #openstack-marconi03:29
*** cpallares has quit IRC03:52
*** haomaiwang has joined #openstack-marconi04:11
*** kgriffs_afk is now known as kgriffs04:13
*** kgriffs is now known as kgriffs_afk04:23
*** haomaiwang has quit IRC04:43
*** haomaiwa_ has joined #openstack-marconi04:44
*** haomaiwa_ has quit IRC04:47
*** haomaiwang has joined #openstack-marconi04:47
*** kgriffs_afk is now known as kgriffs05:14
*** kgriffs is now known as kgriffs_afk05:24
*** reed has quit IRC05:36
*** Abhishek has joined #openstack-marconi05:39
*** Abhishek has quit IRC05:44
*** Abhishek_ has joined #openstack-marconi06:08
*** kgriffs_afk is now known as kgriffs06:15
*** kgriffs is now known as kgriffs_afk06:25
*** flwang has quit IRC06:30
*** flwang has joined #openstack-marconi06:33
*** kgriffs_afk is now known as kgriffs07:16
*** kgriffs is now known as kgriffs_afk07:26
*** jamieh has joined #openstack-marconi07:45
*** Abhishek_ has quit IRC07:46
*** jamieh has quit IRC07:50
*** flaper87|afk is now known as flaper8708:06
*** kgriffs_afk is now known as kgriffs08:17
*** kgriffs is now known as kgriffs_afk08:27
*** rongze has joined #openstack-marconi08:29
*** nosnos has joined #openstack-marconi08:55
*** yassine has joined #openstack-marconi08:58
*** yassine has quit IRC09:00
*** yassine has joined #openstack-marconi09:00
*** yassine has quit IRC09:02
*** rongze has quit IRC09:05
*** yassine has joined #openstack-marconi09:06
*** yassine has quit IRC09:06
*** yassine has joined #openstack-marconi09:07
*** Abhishek_ has joined #openstack-marconi09:14
*** kgriffs_afk is now known as kgriffs09:18
*** kgriffs is now known as kgriffs_afk09:27
*** rongze has joined #openstack-marconi09:34
*** nosnos has quit IRC09:51
*** jamieh has joined #openstack-marconi09:53
*** kgriffs_afk is now known as kgriffs10:18
*** flwang has quit IRC10:25
*** kgriffs is now known as kgriffs_afk10:28
*** flaper87 is now known as flaper87|afk10:33
*** chandankumar has quit IRC10:43
*** chandankumar has joined #openstack-marconi10:46
*** rongze has quit IRC11:06
*** kgriffs_afk is now known as kgriffs11:19
*** kgriffs is now known as kgriffs_afk11:29
*** rongze has joined #openstack-marconi11:38
*** rongze has quit IRC11:49
*** exed has joined #openstack-marconi12:06
*** rongze has joined #openstack-marconi12:15
*** oz_akan_ has joined #openstack-marconi12:18
*** kgriffs_afk is now known as kgriffs12:20
*** malini is now known as malini_afk12:29
*** kgriffs is now known as kgriffs_afk12:30
*** Abhishek_ has quit IRC12:37
*** thomasem has joined #openstack-marconi12:48
*** oz_akan_ has quit IRC12:49
*** kgriffs_afk is now known as kgriffs13:21
*** kgriffs is now known as kgriffs_afk13:31
*** cpallares has joined #openstack-marconi13:42
*** oz_akan_ has joined #openstack-marconi13:51
*** oz_akan_ has quit IRC13:52
*** oz_akan_ has joined #openstack-marconi13:53
*** rongze has quit IRC13:59
*** kgriffs_afk is now known as kgriffs14:22
*** mpanetta has joined #openstack-marconi14:23
*** russellb is now known as rustlebee14:25
*** flaper87|afk is now known as flaper8714:26
*** kgriffs is now known as kgriffs_afk14:32
*** Abhishe__ has joined #openstack-marconi14:41
*** rongze has joined #openstack-marconi14:44
*** rektide has quit IRC14:46
*** oz_akan_ has quit IRC14:46
*** rektide has joined #openstack-marconi14:46
*** oz_akan_ has joined #openstack-marconi14:47
*** malini_afk is now known as malini14:49
*** exed_ has joined #openstack-marconi14:54
*** exed has quit IRC14:56
*** exed_ is now known as exed14:56
*** jcru has joined #openstack-marconi15:00
maliniflaper87: ping15:08
flaper87malini: pong15:08
malinigood afternoon!! how are you?15:08
flaper87malini: I'm doing well, how are you doing?15:09
malinilife would have been better, if I cud figure out devstack </sigh>15:09
*** rongze has quit IRC15:11
*** flwang has joined #openstack-marconi15:20
maliniflaper87: I am running into this , when running devstack15:22
maliniany pointers?15:22
*** kgriffs_afk is now known as kgriffs15:22
*** alcabrera has joined #openstack-marconi15:23
alcabreraGood morning/afternoon/evening! :D15:23
maliniflaper87: this is my oocalrc
malinialcabrera: good morning!!15:23
alcabreramalini: hey! :)15:24
malinihow's your friday starting?15:24
alcabreraawesome. There's banana yogurt, warmth, and lots of rest. Yours?15:24
maliniits going great :)15:25
*** rwsu has joined #openstack-marconi15:31
*** kgriffs is now known as kgriffs_afk15:32
*** kgriffs_afk is now known as kgriffs15:37
* flaper87 back15:38
flaper87alcabrera: good morning15:38
flwangalcabrera: ping15:38
flaper87malini: mmh, that's wierd15:38
flaper87why isn't there a SERVICE_ID set15:39
alcabreraflaper87: hey! :)15:39
alcabreraflwang: pong15:39
*** rongze has joined #openstack-marconi15:39
flwangalcabrera: may I get your attention on my baby ?15:40
alcabreraflaper87: certainly. I'll check it now.15:40
flaper87alcabrera: and you obviously meant flwang15:41
flwangflaper87: morning, superman15:41
flwangalcabrera: thanks15:42
flaper87flwang: gooood morning, buddy!15:42
alcabreraIRC autocomplete got me this time.15:43
alcabreraflwang: +2 - approved! Thanks!15:43
flwangalcabrera: cooooooooooooooooooooooooooool15:44
flwangalcabrera: my 2nd baby on Marconi15:44
alcabrerait's a good morning for merging.15:46
alcabreraask me for reviews while I'm still eating this yogurt. :P15:46
flwangwow, enjoy it15:48
alcabrerathanks! I did. It's all gone now. :)15:50
flaper87pls, thanks!15:52
*** etoews3 has quit IRC15:52
*** etoews2 has joined #openstack-marconi15:52
alcabreraflaper87: even though the yogurt ran out...15:52
alcabreraI'll do it. ;(15:52
alcabrerahahaha, a Nova patch!15:53
alcabreraI'll do it anyway. :P15:53
flaper87just kidding15:53
flaper87no need to waste your time in nova15:53
flaper87you asked for a review, I gave you one15:53
* alcabrera gave up on Nova review, realized couldn't make an informed, fair decision to +1 or -115:55
alcabreraEven for a two line patch15:55
alcabreratoo much EC215:55
maliniflaper87: sorry stepped away15:58
flaper87alcabrera: :P15:59
kgriffsflaper87: now I'm curious16:00
kgriffswhat are some examples of 'volume_type_id''16:00
flaper87kgriffs: just uuids16:01
maliniflaper87: do you know if I need any env variables etc. ?16:01
flaper87kgriffs: or should I say: 8-4-4-4-12 ?16:02
kgriffsflaper87: in EC2 the normal choices are io1 or standard16:02
kgriffsflaper87: LOL16:02
kgriffsI love it when you speak dirty16:02
flaper87kgriffs: yeah, in cinder volume types are custom, though.16:02
alcabrera8-4-4-4-12 is stuck with me forever16:04
mpanettasounds like memory timings from the good ole days16:06
alcabrerampanetta: oh yeah. I remember when I used to look at those numvers and worry about them when shopping on newegg. :P16:06
mpanettahahah yes16:07
alcabrera*numbers, too16:07
mpanettameh, people seem to be using V's to represent all sorts of letters these days, so a B is acceptable :P16:08
*** flwang has quit IRC16:09
openstackgerritA change was merged to openstack/marconi: Return relative URIs for href-template of homedoc
*** rongze has quit IRC16:14
*** rongze has joined #openstack-marconi16:21
*** etoews2 has quit IRC16:37
*** openstackgerrit has quit IRC16:37
*** Alex_Gaynor has quit IRC16:37
*** 17WAAKZAN has joined #openstack-marconi16:38
*** etoews2 has joined #openstack-marconi16:38
*** openstackgerrit has joined #openstack-marconi16:38
*** Alex_Gaynor has joined #openstack-marconi16:38
*** haomaiwang has quit IRC16:39
maliniflaper87: I could get devstack working finally :)16:41
flaper87malini: what was it?16:41
flaper87malini: is there a bug?16:41
maliniHad to make some changes in lib/marconi.sh16:42
malinilet me get you the exact line #16:42
malini - removed 'get_id'16:43
maliniadded | grep ' " id " | get_field 2)' after
maliniI'll submit a patch for this, if it looks ok to you16:44
*** haomaiwang has joined #openstack-marconi16:44
alcabreramalini: awesome!16:47
*** etoews3 has joined #openstack-marconi16:47
*** 17WAAKZAN has quit IRC16:49
*** etoews2 has quit IRC16:49
*** openstackgerrit has quit IRC16:49
*** Alex_Gaynor has quit IRC16:49
*** Alex_Gaynor has joined #openstack-marconi16:54
*** ayoung has joined #openstack-marconi16:56
flaper87malini: mmh, I wonder why that doesn't work. Did you look at other project's sh files?17:06
maliniyes..I looked at neutron17:07
maliniI was using tht as a reference17:07
maliniflaper87: were you able to get it running against ur devstack ?17:07
flaper87malini: if neutron does that grep, then fine!17:08
malinicool..I'll submit a patch for this17:08
*** haomaiwang has quit IRC17:11
*** openstackgerrit has joined #openstack-marconi17:22
*** Alex_Gaynor has quit IRC17:26
*** Alex_Gaynor has joined #openstack-marconi17:26
* flaper87 heads to dinner17:42
*** reed has joined #openstack-marconi17:43
*** exed has quit IRC17:45
maliniyayyy…I just submitted my first non-marconi Openstack patch :)17:47
cpallaresmalini, yay congrats :D17:48
malinithanks cpallares ;)17:48
alcabreramalini: sweet! I've yet to contribute outside of marconi, iirc. :)17:49
maliniIt feels like I have stepped out into the big bad world :D17:49
alcabreramalini: it's pretty nice out there - most of the time. ;)17:57
alcabrerabefore too long, you'll find yourself submitting patches to17:57
alcabrerayou'll have touched it all. :D17:57
malinitempest is coming soon17:57
malinitht's my top priority now17:58
alcabreragood luck!17:59
*** alcabrera is now known as alcabrera|afk18:14
*** rwsu has quit IRC18:32
*** jergerber has joined #openstack-marconi18:35
*** rwsu has joined #openstack-marconi18:40
*** alcabrera|afk is now known as alcabrera18:59
cpallareswhere are the files that deal with server connection?19:12
cpallares...or are they spread out?19:12
cpallaresi'm looking into this bug
*** rongze has quit IRC19:20
*** yassine has quit IRC19:30
alcabreracpallares: hmmm19:31
alcabreracpallares: the places to look should all be in marconi/queues/transport19:32
alcabrerafor server connections, WSGI/HTTP style, marconi.queues.transport.wsgi.{app,driver} are especially relevant.19:32
cpallaresalcabrera: thanks!19:32
alcabrerafor handling specific requests, they19:32
alcabrera're the rest of the file sin transport/wsgi/*.py19:33
alcabreralike,, etc.19:33
alcabreranp. :)19:33
cpallaresalcabrera: i was looking in transport but there's a great deal of errors that aren't logged so I wasn't sure which ones are the server ones19:33
alcabreraI'd say that whatever would result in an HTTP 500 is a server error.19:35
alcabreraI think there's actually a candidate sitting in queues/storage/sharding.py19:35
alcabreraI'm not sure if when kgriffs and I got together and implemented that file, if we logged exceptional conditions.19:35
cpallaresalcabrera: hmm, what causes an HTTP 500 error? Is that any storage error?20:04
* cpallares doesn't know anything about servers or server errors20:04
alcabreracpallares: for the case of marconi, storage errors are a big part of it. If we try to communicate with storage and it fails -> 50020:05
alcabreraif no shards are registered in a sharded marconi context -> 50020:05
alcabreralemme see20:05
alcabreraif we try to contact a shard that is registered with the wrong URL or is down -> 50020:06
kgriffsflaper87: ETA on a new cache patch?20:08
mpanettaOh yay, someone is looking at my bug. :)20:08
alcabreracpallares: That's all the cases I can think of at the moment.20:09
cpallaresthank you alcabrera, it gives me a start20:09
mpanettaIf the catalog DB is down...20:10
alcabreraAny operation that communicate with storage or over the network are suspect for 500s. :)20:10
mpanettaPretty much20:10
mpanettaI am trying to think if we hit anything else weird20:11
mpanettaWhat about memcached access alcabrera?  Does the code do anything to handle a down memcached server?20:11
alcabrerampanetta: memcached would do it, too. Good thought!20:11
alcabreraYeah, I don't think we handle caching errors, either20:12
alcabreraIn that case, I'd like to see an error logged, but just continue operating in a degraded state20:12
mpanettaI know when a shard does not exist and you are trying to create a queue it throws an odd exception in the logs, something along the lins of ShardNotFound does not exist...20:12
mpanettaBut nothing else is logged.20:12
kgriffsmpanetta: btw, in your latest marconi.conf file did you remove the governor options per my comment in the test config?20:13
mpanettaAnd kept your comment20:13
kgriffskewl, thanks. That should help make rate limiting work better20:13
mpanettano problem :)20:14
*** Abhishe__ has quit IRC20:15
openstackgerritKurt Griffiths proposed a change to openstack/marconi: chore: Bump version to icehouse-2 dev
kgriffsalcabrera: the python memcached driver fails silently when it can't reach a node iirc20:30
*** rongze has joined #openstack-marconi20:31
kgriffsI think it pretends to succeed or something. I can't remember exactly. :p20:31
alcabreraIf it returns None, it's perfect for me. :P20:32
alcabreralookup :: Cache -> a -> Maybe b20:32
alcabreraI'll check out oslo.cache failure semantics later.20:32
alcabreraThat'll go on my TODO20:33
kgriffsalcabrera: oslo.cache may not be able to help it, if the underlying library doesn't tell it anything went wrong20:35
kgriffsbut yeah, something to follow up on20:36
kgriffswe need errors to be detected and surfaced to the app, then dealt with correctly there20:36
*** rongze has quit IRC20:36
kgriffsalcabrera: question20:42
kgriffsis u'fsfa' ok, or is size.text_type('ffo') preferred for string literals?20:42
mpanettaalcabrera: what do you think about controlling depolyment with an array of these and then one of these:
mpanettalol it would be like setting off a missile in a military base20:44
*** cpallares has quit IRC20:45
alcabrerakgriffs: u'fsfa' is okay, as long we don't care to support py3.2 (which is fine, IMO)20:45
*** cpallares has joined #openstack-marconi20:45
kgriffsmpanetta: nice. I added that to my wishlist20:45
kgriffsalcabrera: kk20:45
alcabrerab'a' also works across py26, py27, and py3320:46
alcabrerathe dangerous one is 'a'20:46
alcabrerawhich means unicode in py3 and byte-like in py220:46
alcabrerakgriffs: lol - now launching marconi...20:47
kgriffsalcabrera: is there a "unicode" magic method?20:47
kgriffsI know there is __str__20:47
alcabrerahmm, that's new to me. ^20:48
alcabrerakgriffs: I'm not sure __unicode__ is canonical. Lemme check the magic method docs.20:49
alcabreraIt isn't listed when I dir() a new class.20:49
kgriffsI saw it here:
kgriffsquestion is, what is the correct thing to do for py3 compat?20:50
alcabrerakgriffs: it's legit -
kgriffsin py3, __str__ is called for str(mything)20:50
kgriffsand unicode(mything) doesn't work20:50
alcabrerain python 3, NameError: name 'unicode' is not defined20:50
alcabreraso the correct thing, I think, is to avoid __unicode__ entirely.20:51
alcabreraand treat __str__ as if it should return a unicode display20:51
mpanettaOk, I think I am going to have to make a button->queue thingy20:52
kgriffsalcabrera: if __unicode__ is not defined, then __str__ is called and the result is coerced to unicode20:53
mpanettaare there python bindings to linux HID?  Anyone know off hand?20:53
kgriffsalcabrera: I don't like the sound of "coerced" - could mess things up if the wrong encoding is assumed20:53
alcabrerakgriffs: hmmm... it certainly sounds tricky.20:54
alcabrera__str__ magic makes me nervous20:54
kgriffsI don't think this is done correctly, and am trying to work out a better approach20:56
kgriffsmay have to have an if six.PY320:56
kgriffsI was thinking, first of all, that message and msg_fmt should be defined to be wide20:56
kgriffsunder py3 you just return the result directly from __str__ and __unicode__ is not needed20:57
kgriffsunder py2 you encode with utf8 in __str__ and you return directly in __unicode__20:57
kgriffsalcabrera: thoughts on that approach?20:57
alcabreraI'm +1 for making message and msg_fmt wide20:58
alcabreraThat'll help things along.20:58
alcabreraregarding the return mechanism20:59
alcabreraI'm not sure, but I'm pondering on it20:59
alcabreraI wonder what kind of weird things would be returned by u'é'.encode('utf8') under __str__ in python 2...?21:01
alcabreraI like __unicode__ just passing through21:01
alcabreraand __str__ just passing through in py321:01
alcabrerait's that __str__ in py2 that makes me nervous21:01
alcabreraI'm leaning towards agreeing with regards to using if PY321:01
alcabrerakgriffs: ^^21:02
alcabrerau'é'.encode('utf8') -> '\xc3\xa9'21:02
alcabreraso that might not be... desirable. :/21:02
alcabrerait seems like __str__ should always be readble.21:02
alcabreraso my best guess is that the right choice might just be to pass through the __unicode__ in __str__ for py2.21:03
alcabrerasince the way forward depends on __str__ exposing a unicode string21:03
alcabrera /endthoughts21:04
*** malini is now known as malini_afk21:04
kgriffswoah hold on21:16
kgriffsalcabrera: what do you think of this -
* alcabrera checks21:20
alcabrerakgriffs: I'm concerned about  return self.message.encode('utf-8')21:21
alcabrerasince if message contains any non-ascii characters, it's going to show up as an unreadable mess when displayed to the user.21:22
kgriffswhat is the alternative?21:22
kgriffsinsert question marks?21:22
alcabreraJust overload __str__ to return a unicode string21:22
alcabreraI think the coercion only happens if __str__ is not overloaded.21:23
kgriffsI tried it21:23
kgriffsif you return unicode you get21:23
kgriffsUnicodeDecodeError: 'ascii' codec can't decode byte 0xe2 in position 0: ordinal not in range(128)21:23
alcabrerawell then...21:23
kgriffsideally, you would just always use six.text_type(ex)21:24
alcabreraThough that will raise UnicodeDecodeError...21:25
kgriffsyes, that is what the interpreter does automatically when it coerces21:26
kgriffsidk if it is possible to tell it to use a different codec by default21:26
alcabreraidk, either.21:27
kgriffsalcabrera: here is my final draft:
kgriffslook OK?21:28
alcabrerakgriffs: yup21:28
alcabreraI think that's the safest bet21:28
alcabreraNeeds to come with strong dissuasion against str().21:29
alcabrerasix.text_type is best21:29
alcabreraI'm heading out for the day.21:29
alcabreraHave a good weekend, everyone. :)21:29
*** alcabrera has quit IRC21:30
*** oz_akan_ has quit IRC22:08
*** thomasem has quit IRC22:11
*** mpanetta has quit IRC22:15
*** cpallares has quit IRC22:33
*** jamieh has quit IRC22:35
*** jamieh has joined #openstack-marconi22:46
*** kgriffs is now known as kgriffs_afk22:50
*** ametts has quit IRC23:16
*** oz_akan_ has joined #openstack-marconi23:19
*** oz_akan_ has quit IRC23:23
*** flaper87 is now known as flaper87|afk23:50
*** jcru has quit IRC23:53

Generated by 2.14.0 by Marius Gedminas - find it at!