Wednesday, 2014-10-22

*** vkmc has quit IRC00:20
*** reed has quit IRC00:20
*** jeffrey4l has quit IRC00:20
*** zhiyan has quit IRC00:20
*** boris-42 has quit IRC00:20
*** vipul has quit IRC00:20
*** X019 has quit IRC00:20
*** cpallares has quit IRC00:20
*** yfujioka has quit IRC00:20
*** ekarlso has quit IRC00:20
*** flwang1 has quit IRC00:20
*** openstackgerrit has quit IRC00:20
*** wpf has quit IRC00:20
*** flwang has quit IRC00:20
*** dmitryme has quit IRC00:20
*** AAzza_afk has quit IRC00:20
*** sebasmagri has quit IRC00:20
*** flaper87 has quit IRC00:20
*** tmu has quit IRC00:20
*** notmyname has quit IRC00:20
*** riveter has quit IRC00:20
*** dguerri has quit IRC00:20
*** malini|afk has quit IRC00:20
*** kgriffs|afk has quit IRC00:20
*** X019 has joined #openstack-zaqar00:26
*** boris-42 has joined #openstack-zaqar00:26
*** flwang1 has joined #openstack-zaqar00:26
*** reed has joined #openstack-zaqar00:26
*** malini|afk has joined #openstack-zaqar00:26
*** jeffrey4l has joined #openstack-zaqar00:26
*** openstackgerrit has joined #openstack-zaqar00:26
*** yfujioka has joined #openstack-zaqar00:26
*** flwang has joined #openstack-zaqar00:26
*** wpf has joined #openstack-zaqar00:26
*** dmitryme has joined #openstack-zaqar00:26
*** ekarlso has joined #openstack-zaqar00:26
*** AAzza_afk has joined #openstack-zaqar00:26
*** zhiyan has joined #openstack-zaqar00:26
*** sebasmagri has joined #openstack-zaqar00:26
*** riveter has joined #openstack-zaqar00:26
*** flaper87 has joined #openstack-zaqar00:26
*** vipul has joined #openstack-zaqar00:26
*** tmu has joined #openstack-zaqar00:26
*** kgriffs|afk has joined #openstack-zaqar00:26
*** dguerri has joined #openstack-zaqar00:26
*** notmyname has joined #openstack-zaqar00:26
openstackgerritRose Ames proposed a change to openstack/zaqar: Remove duplicated code in pooling meta-controllers  https://review.openstack.org/12910901:53
openstackgerritJeffrey Zhang proposed a change to openstack/zaqar: Version discovery for root URI  https://review.openstack.org/13009402:14
wpfjeffrey4l:  ping02:18
jeffrey4lwpf, pong02:18
wpfyou are the Jeffrey Zhang ,right ?02:18
jeffrey4lyup02:19
jeffrey4lsup?02:19
wpfjeffrey4l:  :p  ,  Ok ,I saw your comments for my PS ,about the x-openstack-request-id logline, thanks02:19
jeffrey4li remember that.02:20
wpfjeffrey4l:  anyway,  do you know what how the 'request_id' context was plugged in ?02:21
wpfI try to find it ,but did not get the entry point02:21
jeffrey4lwpf, do u mean how the request_id is generated in the context object?02:23
wpfand I debug the zaqar code,  r record.__dict__.get('request_id')  is None02:24
jeffrey4lyep. I know that.02:24
jeffrey4lHere is the magic.02:24
wpfthanks02:24
* jeffrey4l searching the code02:24
jeffrey4lbtw, this bp is needed for this feature https://blueprints.launchpad.net/zaqar/+spec/log-project-id02:25
jeffrey4lwpf, check this line https://git.openstack.org/cgit/openstack/zaqar/tree/zaqar/openstack/common/log.py#n31602:26
wpfI thought this bp had not been implemented02:26
jeffrey4lwpf, yes. I am trying to implement the bp.02:26
jeffrey4lthe context object is load from `local.store` ( a threading local weak object) https://git.openstack.org/cgit/openstack/zaqar/tree/zaqar/openstack/common/local.py#n3802:27
wpfjeffrey4l:   you know , in my PS,  I only generate the uuid and set it into the req.env and also the response.headers using the before/after handler of Falcon,02:28
jeffrey4lwpf, Now, there is no code to set the local.store variable.02:29
wpfso do you mean that I need to put the request-id into the context?02:29
jeffrey4lwpf, in your PS, my comments mean that why the request-id should be exist when and only the status_code < 40002:30
jeffrey4lhow about when the server raise 404/500 error, i think the request-id should also be exist.02:31
jeffrey4lI think putting the request-id into the context is another things.02:31
wpfit depends I thought02:31
wpfsince there are also some other validations before the request was really handled, such as if there is no CLIENT-ID,02:32
wpfthen the request will be refused by 400 or some other exception02:33
wpfin that case , I don't think we need the request-id02:33
wpfand of course we may put the 'generated request id ' as the first step for any requests, but do we really need that?02:34
jeffrey4lwhy? reqest-id can help us debug the code.  I'd like the request_id exist in all situation.02:34
wpfOk,  I will think it again , (actually ,I have not read you inline comments, just saw that you give me the clue of the log info , -:))02:36
jeffrey4lroger that. ;p02:37
wpfbut I still have no idea about how to put the request-id into the context02:38
wpf-____-02:38
jeffrey4lwpf, aha02:40
jeffrey4lit should be put in the loca.store in the beginning of the request.02:40
jeffrey4lwpf, pls check the nova context code http://git.openstack.org/cgit/openstack/nova/tree/nova/context.py#n97 . when the context is created, it will be store into local.store( a threading weakref objects)02:45
jeffrey4lthen when logging, the context can be accessed by the logging module.02:45
jeffrey4lthe request-id is include in the context objects.02:45
wpfOk, thanks , I will try to understand it ,02:45
jeffrey4lhttp://git.openstack.org/cgit/openstack/nova/tree/nova/context.py#n7402:46
wpfand will ping you again if there are some questions , :P02:46
jeffrey4lno problem.02:46
jeffrey4lI think you patch should focus on the request-id in the environment. and my incoming patch( implements https://blueprints.launchpad.net/zaqar/+spec/log-project-id) should encapsulation is into context object and logging it every where.02:47
wpf-:)02:48
*** flwang1 has quit IRC03:34
*** reed has quit IRC03:45
*** fifieldt has joined #openstack-zaqar04:08
openstackgerritJeffrey Zhang proposed a change to openstack/python-zaqarclient: Fix the client name typo  https://review.openstack.org/13011604:22
*** sgotliv_ has joined #openstack-zaqar05:21
wpfjeffrey4l:  ping05:28
jeffrey4lwpf, pong05:34
wpfjeffrey4l:   when on_xxxxxx(such as on_post) raise some exception , the 'after' hook will be executed in Falcon05:35
wpfI thought I can't make sure that all >400 responses had the header since some of the 400 response are exceptions ,and some of them are normal response just with status-code is 40005:36
jeffrey4lwpf, yup. I need check the Falcon's code to make sure of that.05:38
wpf@wraps(responder)05:39
wpf    def do_after(req, resp, **kwargs):05:39
wpf        responder(req, resp, **kwargs)05:39
wpf        action(req, resp)05:39
openstackgerritPeng Fei Wang proposed a change to openstack/zaqar: Add x-openstack-request-id to zaqar responses  https://review.openstack.org/12881505:45
jeffrey4lwpf, I checked the code and know the reason.06:21
wpfjeffrey4l:  yep06:22
jeffrey4lBut i think still think the request-id should be exist all the time, no matter whether the exception is raised.06:22
jeffrey4lwpf, So i think it should be a bug/feature( falcon should handle the exception more gracefully). Or we can use middleware to handle the exception.06:24
wpfOk, when flaper87 who worked on Falcon I thought is online , I will get his insights06:26
jeffrey4lwpf, another question: I doubt that whether the  request id should be created every  time and never passed to other project( i.e. keystone). IMHO, the cross-service-request-id mean that in one request through several service, the request-id should keep the same.06:29
jeffrey4lI don't know whether i understand correctly.06:30
jeffrey4lif so, there is more work to do. For example, pass the request id to keystone when necessarily.06:31
jeffrey4lBut it seems that the cross-service-request-id is not implemented in keystone. :(06:31
wpfyes, no x-openstack-request-id in keystone06:33
* wpf dig a big hole for my self06:33
jeffrey4lwpf, pls check https://blueprints.launchpad.net/openstack/?searchtext=cross-service-request-id and https://review.openstack.org/#/c/70823/06:33
jeffrey4lwpf, LOL06:33
*** sgotliv__ has joined #openstack-zaqar06:33
jeffrey4lThe request id does not involve Keystone, nor should it be generated by Keystone: it should be generated by Heat and or Nova when scheduling a long running task. Keystone is not part of that.06:35
jeffrey4lcomments from https://review.openstack.org/#/c/70823/06:35
wpfthen why our zaqar need it ?  :P06:36
wpfit should should be pretty fast tasks within zaqar06:36
*** sgotliv_ has quit IRC06:37
jeffrey4lat least, zaqar need connect to keystone. In this case, cross request id is needed, i think.06:37
*** sgotliv__ has quit IRC06:38
wpfok, it's cross-service06:39
jeffrey4li think so.06:39
*** X019 has quit IRC06:39
openstackgerritJeffrey Zhang proposed a change to openstack/zaqar: Always include the project id in the logs  https://review.openstack.org/13013907:01
*** flwang1 has joined #openstack-zaqar07:08
wpfflaper87:  ping07:22
flaper87wpf: pong07:24
wpfflaper87:  one question in Falcon07:25
wpfif the responder raises some exceptions ,then the after hook will not be executed ,right ?07:26
wpfjeffrey4l:  I saw that there always be one request-id in the request context when injecting,  can we add it into the response header?07:30
flaper87wpf: mmh, I think that's right, yes! I mean, the after hooks shouldn't be07:30
wpfflaper87:  Ok, since jeffrey4l  want to all the response (even raise exception) have to the request_id,  so with after hook, it will not work07:31
jeffrey4lwpf, I think this should be different thing with cross-service-request-id.07:31
*** exploreshaifali has joined #openstack-zaqar07:32
flaper87wpf: I should probably double check that07:32
flaper87but it'd be easier to just try it out :P07:32
wpfI'd check the Falcon code, just confirm with you :P07:32
jeffrey4lIf the cross service request id is what I understand, there should be more work to do( discussed above)07:32
wpfjeffrey4l:  you mean transfer it to keystone ?07:33
jeffrey4lwpf, right.07:33
*** AAzzal has joined #openstack-zaqar07:33
*** AAzzal is now known as AAzza07:33
wpfjeffrey4l:  Ok ,actually when I check nova related code,  it seems that they only add the request-id into request.env and response header with middleware, did not something with the keystone07:34
wpfmaybe I need to check it again.07:34
*** AAzzal has joined #openstack-zaqar07:34
jeffrey4lso I want separate those two things and don't want to add it into the response header. Even though it it easy.07:35
*** AAzza_afk has quit IRC07:36
jeffrey4lwpf, About the cross-service-request-id, I think that none of the project/service implement it. So we find nothing in the nova's code.07:36
jeffrey4lwpf, check this. https://blueprints.launchpad.net/openstack/?searchtext=cross-service-request-id none service implement it.07:37
wpfyes, I agree that it's two things ,but it should not prevent us to add it into the response header I though,  since in my project , I usually add and id into the request log (such as xxx-id >> sina.com) , and then add the same id into the response log line (such as xxx-id << 200 Ok)07:37
*** AAzza has quit IRC07:37
*** AAzzal is now known as AAzza07:37
wpfwhich will make the debug very easy.07:38
wpfso I don't get your point that why you did not want to add it into the response header.07:38
jeffrey4lyou mean add the response id into the logs or headers?07:39
wpfrequest-id07:39
jeffrey4lso , you mean add the request-id into the headers?07:40
wpfyes07:40
wpfI am just curious that why we do not , it seems to be nothing related with the cross-service07:41
jeffrey4lI am trying the clear your question. I think there is some misunderstand between us.07:41
jeffrey4lAfter that, you can log the request-id( read from headers) into logs?07:41
jeffrey4lthis is what u want? right?07:41
wpfno, with your changes,  the request-id is already in the log ,right?07:42
jeffrey4lyep07:42
jeffrey4lSo why it is necessary to add it to the response header?07:43
wpfI thought it's easy for user to find out that which lines in the logs are for his request07:44
jeffrey4llogs include the request id and the line number. It isn't solve the problem?07:45
*** AAzza is now known as AAzza_afk07:46
wpfI am not sure how the request_id was logged,  is it a pair in the log file,  the first line is the entry of request ,and then blablabla,,,  then another line with the same id for the response ?07:47
*** X019 has joined #openstack-zaqar07:47
jeffrey4lyes.07:48
jeffrey4lwpf, something like below07:49
jeffrey4l2014-10-22 15:48:56.115 24496 DEBUG zaqar.queues.transport.wsgi.v1_0.queues [req-b3f3436b-cb66-46db-852f-d91550b839c2 - - - - -] [project_id:a] Queue collection GET on_get /home/jeffrey/openstack/zaqar/zaqar/queues/transport/wsgi/v1_0/queues.py:9107:49
jeffrey4lreq-xxxx is the request id and it will exist in all the logs.07:49
wpfOk, suppose that I make the call with curl or some other client, as at the same time ,maybe other also make the same request with different body,  I am not sure whether I need the response header to find 'my' log in the log file or not07:51
wpf:-)07:51
wpfmaybe you can also output the client-id07:52
jeffrey4lwpf, so when u want the log for a certain request, u can filter the log with a certain request-id. U can get all the related log .07:52
wpfyes07:52
jeffrey4lwpf, the client-id is optional.07:52
wpfsince I don't know which line is for me07:52
wpfoptional?07:53
jeffrey4lwpf, u can modify the logging_context_format_string conf to add the client-id.07:53
jeffrey4lwpf, wrong word. Should use configurable.07:53
wpfok07:53
jeffrey4llogging_context_format_string=%(client_id)s07:54
jeffrey4lsome thing like above.07:54
wpfyes, got it07:54
jeffrey4lall the value in the context variable can be logged through above method.07:54
jeffrey4l:p07:54
wpfanyway , seems that I can abandon my PS, :)07:54
jeffrey4lI think so. We can implement this until the cross-service-request-id implemented in keystone at least.07:56
*** sgotliv__ has joined #openstack-zaqar07:59
*** sgotliv__ has quit IRC07:59
*** sgotliv has joined #openstack-zaqar07:59
wpfjeffrey4l:  ok, thanks for your help .   I need to understand the magic for the log context08:01
jeffrey4lwpf, :D08:02
*** AAzza_afk is now known as AAzza08:10
*** sgotliv has quit IRC08:46
*** sgotliv has joined #openstack-zaqar08:49
*** yfujioka has quit IRC09:11
*** flwang1 has quit IRC09:12
*** flwang1 has joined #openstack-zaqar09:12
*** jeffrey4l has quit IRC09:19
*** exploreshaifali has quit IRC09:28
*** sgotliv has quit IRC09:29
*** exploreshaifali has joined #openstack-zaqar09:29
*** jeffrey4l has joined #openstack-zaqar09:32
*** sgotliv has joined #openstack-zaqar09:41
*** bradjones has joined #openstack-zaqar09:58
*** bradjones has quit IRC10:08
*** bradjones has joined #openstack-zaqar10:08
*** flwang2 has joined #openstack-zaqar10:13
*** flwang1 has quit IRC10:13
*** sgotliv has quit IRC10:18
*** exploreshaifali has quit IRC10:45
*** sgotliv has joined #openstack-zaqar11:00
*** exploreshaifali has joined #openstack-zaqar11:03
*** sgotliv has quit IRC11:10
*** malini has joined #openstack-zaqar11:11
*** sgotliv has joined #openstack-zaqar11:13
*** exploreshaifali has quit IRC11:24
*** bradjones has quit IRC11:30
*** bradjones has joined #openstack-zaqar11:32
*** sgotliv has quit IRC11:41
*** sgotliv has joined #openstack-zaqar11:42
*** exploreshaifali has joined #openstack-zaqar11:44
*** vkmc has joined #openstack-zaqar12:32
vkmcmorniiiing12:33
*** sgotliv has quit IRC12:38
*** sgotliv has joined #openstack-zaqar12:51
*** sgotliv has quit IRC12:56
*** sgotliv has joined #openstack-zaqar13:02
exploreshaifaliflaper87, around?13:02
exploreshaifaliflaper87, what should I do now?13:04
*** jchai has joined #openstack-zaqar13:05
*** sriram has joined #openstack-zaqar13:06
*** malini has quit IRC13:06
flaper87exploreshaifali: yo yo13:10
flaper87back13:10
flaper87exploreshaifali: what do you mean?13:10
flaper87vkmc: good morning, young lady!13:10
*** mpanetta has joined #openstack-zaqar13:12
exploreshaifaliflaper87, I have submitted the application, now how to proceed ahead?13:16
vkmcflaper87, morning good sir13:17
exploreshaifalivkmc, I need one suggestion13:18
exploreshaifaliI have put details and schedules for 3 projects in application13:18
exploreshaifalivkmc, I was wondering if it will make my application weaker by showing interest in many projects13:19
exploreshaifalivkmc, what do you think?13:19
flaper87exploreshaifali: next step is to wait for the approval13:19
exploreshaifaliflaper87, but till then I think I should try out to fix any bug for zaqar as I haven't contributed to it?13:20
flaper87exploreshaifali: that sounds perfect to me and it's definitely a great idea13:21
flaper87:D13:21
exploreshaifalior any other stuff like so13:21
flaper87exploreshaifali: have you looked into the bug list?13:21
exploreshaifaliyup13:21
exploreshaifalithere is no any low hanging fruit bug13:21
exploreshaifalican you suggest any bug which will be easy to fix for someone new to zaqar?13:22
exploreshaifalior I there is not, I think I should try for any bug that I find intutive as there is loong time in  approval date13:23
flaper87exploreshaifali: the later, my advice is to always jump in, read and take a look at what feels good13:24
flaper87not necessarily easy, it may also be a difficult bug13:24
flaper87you just need to be interested in working on it13:24
exploreshaifaliyeah13:24
exploreshaifaliflaper87, have you seen https://www.youtube.com/watch?v=seUzASB8sxU13:25
exploreshaifali:P13:25
*** flwang2 has quit IRC13:25
*** flwang1 has joined #openstack-zaqar13:26
flaper87exploreshaifali: LOOOOOOOOOL, I hadn't13:34
exploreshaifali:D13:34
*** exploreshaifali has quit IRC13:39
*** malini has joined #openstack-zaqar13:46
*** ametts has joined #openstack-zaqar13:47
*** cpallares has joined #openstack-zaqar13:51
*** AAzza has quit IRC13:57
*** amitgandhinz has joined #openstack-zaqar13:58
*** AAzza_afk has joined #openstack-zaqar13:59
*** AAzza_afk is now known as AAzza13:59
*** exploreshaifali has joined #openstack-zaqar14:01
*** jchai is now known as jchai_afk14:11
*** vkmc has quit IRC14:13
*** jchai_afk is now known as jchai14:25
*** vkmc has joined #openstack-zaqar14:31
rivetermornin vkmc14:43
vkmcmorning riveter!14:45
*** jeffrey4l has quit IRC14:52
vkmcflaper87, around? can you review this one https://review.openstack.org/#/c/129988/?15:01
flaper87vkmc: around now15:08
*** reed has joined #openstack-zaqar15:10
*** kgriffs|afk is now known as kgriffs15:13
*** exploreshaifali has quit IRC15:17
*** vkmc has quit IRC15:17
flaper87vkmc left >.>15:17
*** vkmc has joined #openstack-zaqar15:24
*** vkmc has joined #openstack-zaqar15:24
flaper87vkmc: stop running away from me15:24
* flaper87 calls his buddy: THE PENGUIN!15:25
flaper87vkmc: I left a comment :)15:25
vkmcflaper87, heh sorry, the connection is not very good here15:25
vkmcI'm in the conference15:25
vkmc:T15:25
flaper87vkmc: no no no, do not play the conference card on me... NO!15:25
flaper87I refuse to be treated like that15:25
* flaper87 is calling his lawyer15:25
vkmchahaha15:25
* flaper87 will sue vkmc15:26
riveteris your lawyer a penguin?15:26
vkmcnooo, I promise I won't run away again15:26
flaper87riveter: no no, the penguin is THE PENGUIN15:29
vkmcda penguin15:30
riveterbut does THE PENGUIN or does THE PENGUIN not have a law degree?15:31
riveterI suspect you have not even bothered to ascertain this basic information15:31
kgriffsflaper87: you've seen the original of that gummy bear song, right?15:31
flaper87kgriffs: no15:37
flaper87:P15:37
flaper87riveter: exactly, it just gets the job done15:38
flaper87but we can just say it has one15:38
* kgriffs drops a link https://www.youtube.com/watch?v=VFM-_ncS8ZY15:39
flaper87LOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOL15:40
flaper87ok, that's it!15:40
*** exploreshaifali has joined #openstack-zaqar16:02
riveterhi explore16:02
exploreshaifalihi riveter16:03
*** bradjones has quit IRC16:43
*** jchai is now known as jchai_afk17:00
*** jchai_afk is now known as jchai17:06
vkmcflaper87, could you give me more context on the comment you left in https://review.openstack.org/12998817:11
vkmc_17:11
*** cpallares has quit IRC17:38
kgriffsflaper87, riveter: when you both have a minute, I'd like to discuss https://review.openstack.org/#/c/129109/17:47
riveterI'm here17:47
kgriffsflaper87: around?17:47
riveterdo you want to talk now or should we wait til he's around?17:54
*** sgotliv has quit IRC18:07
riveterkgriffs:  I'm going out in half an hour or so but I'll be on later18:09
kgriffsok18:09
kgriffswhen you get back lets try to ping flaper87 again18:10
riveterI could talk now, just thought I mention that I'd be going18:10
riveterso he didn't show up and then I jump up and leave :)18:10
riveterflaper87, around?18:11
riveterok, I'm heading out, will ping you when I get back18:39
*** AAzza is now known as AAzza_afk18:51
*** exploreshaifali has quit IRC19:04
*** amitgandhinz has quit IRC19:13
*** AAzza_afk is now known as AAzza19:18
*** amitgandhinz has joined #openstack-zaqar19:24
*** malini has quit IRC19:30
*** kgriffs is now known as kgriffs|afk19:44
*** flwang1 has quit IRC19:48
*** AAzza is now known as AAzza_afk19:51
*** exploreshaifali has joined #openstack-zaqar19:57
*** AAzza_afk is now known as AAzza20:15
*** kgriffs|afk is now known as kgriffs20:17
*** AAzza is now known as AAzza_afk20:18
*** sgotliv has joined #openstack-zaqar20:32
*** AAzza_afk is now known as AAzza20:42
*** AAzza is now known as AAzza_afk20:44
*** exploreshaifali has quit IRC20:50
*** vkmc has quit IRC20:53
*** X019 has quit IRC20:59
*** sgotliv has quit IRC20:59
*** boris-42 has quit IRC21:00
*** malini has joined #openstack-zaqar21:03
*** sriram has quit IRC21:04
*** malini has quit IRC21:04
*** malini has joined #openstack-zaqar21:05
*** vkmc has joined #openstack-zaqar21:06
*** boris-42 has joined #openstack-zaqar21:07
*** vkmc has quit IRC21:12
*** mpanetta has quit IRC21:14
flwangflaper87: do we have another meeting to discuss the topic of summit?21:14
*** bradjones has joined #openstack-zaqar21:17
*** flwang1 has joined #openstack-zaqar21:54
*** vkmc has joined #openstack-zaqar21:57
*** vkmc has joined #openstack-zaqar21:57
*** amitgandhinz has quit IRC21:58
*** amitgandhinz has joined #openstack-zaqar22:00
*** amitgandhinz has quit IRC22:01
*** malini has quit IRC22:04
*** jchai has quit IRC22:08
*** vkmc has quit IRC22:15
*** amitgandhinz has joined #openstack-zaqar23:12
riveterkgriffs, ping23:16
*** amitgandhinz has quit IRC23:17
*** kgriffs is now known as kgriffs|afk23:39
*** amitgandhinz has joined #openstack-zaqar23:43

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!