*** catintheroof has quit IRC | 00:03 | |
*** catintheroof has joined #openstack-glance | 00:11 | |
*** markvoelker has quit IRC | 00:11 | |
*** catintheroof has quit IRC | 00:25 | |
*** zhurong has joined #openstack-glance | 00:44 | |
*** tovin07_ has joined #openstack-glance | 00:48 | |
*** lucasxu has joined #openstack-glance | 00:49 | |
*** boris-42_ has quit IRC | 00:59 | |
*** mfedosin has quit IRC | 01:12 | |
*** markvoelker has joined #openstack-glance | 01:36 | |
*** lucasxu has quit IRC | 01:37 | |
*** mfedosin has joined #openstack-glance | 01:45 | |
*** mingyu has joined #openstack-glance | 01:55 | |
*** mingyu has quit IRC | 02:04 | |
*** bkopilov has quit IRC | 02:06 | |
*** zhurong has quit IRC | 02:07 | |
openstackgerrit | KaiLi proposed openstack/glance master: Add linked clone option in glance metadata https://review.openstack.org/468327 | 02:16 |
---|---|---|
*** nicolasbock has joined #openstack-glance | 02:17 | |
*** chlong has quit IRC | 02:42 | |
*** mingyu has joined #openstack-glance | 03:01 | |
*** mfedosin has quit IRC | 03:06 | |
*** mingyu has quit IRC | 03:38 | |
*** mingyu has joined #openstack-glance | 03:38 | |
*** mingyu has joined #openstack-glance | 03:44 | |
*** zhurong has joined #openstack-glance | 03:50 | |
*** gcb has quit IRC | 03:58 | |
*** bkopilov has joined #openstack-glance | 04:00 | |
*** mingyu has quit IRC | 04:05 | |
*** mingyu has joined #openstack-glance | 04:10 | |
*** ratailor has joined #openstack-glance | 04:10 | |
*** mingyu has quit IRC | 04:15 | |
*** zhurong has quit IRC | 04:27 | |
*** links has joined #openstack-glance | 04:32 | |
*** zhurong has joined #openstack-glance | 04:33 | |
*** adisky__ has joined #openstack-glance | 04:42 | |
*** bkopilov has quit IRC | 04:43 | |
*** gyee has quit IRC | 04:47 | |
*** udesale has joined #openstack-glance | 04:52 | |
*** mabrams has joined #openstack-glance | 05:05 | |
*** bkopilov has joined #openstack-glance | 05:10 | |
*** bkopilov has quit IRC | 05:23 | |
*** gcb has joined #openstack-glance | 05:30 | |
*** efried has quit IRC | 05:34 | |
*** tshefi has joined #openstack-glance | 05:36 | |
*** bkopilov has joined #openstack-glance | 05:36 | |
*** aavraham has joined #openstack-glance | 05:37 | |
*** gcb has quit IRC | 05:46 | |
*** gcb has joined #openstack-glance | 05:46 | |
*** efried has joined #openstack-glance | 05:50 | |
*** mingyu has joined #openstack-glance | 05:56 | |
*** groen692 has joined #openstack-glance | 05:57 | |
*** rcernin has joined #openstack-glance | 06:05 | |
*** pcaruana has joined #openstack-glance | 06:06 | |
*** mingyu has quit IRC | 06:07 | |
*** mingyu has joined #openstack-glance | 06:09 | |
*** mingyu has quit IRC | 06:23 | |
*** markvoelker has quit IRC | 06:25 | |
*** mabrams has quit IRC | 06:27 | |
*** mabrams has joined #openstack-glance | 06:28 | |
*** mingyu has joined #openstack-glance | 06:28 | |
*** mabrams has quit IRC | 06:28 | |
*** mine0901 has joined #openstack-glance | 06:31 | |
*** mabrams has joined #openstack-glance | 06:33 | |
*** tesseract has joined #openstack-glance | 06:41 | |
*** namnh has joined #openstack-glance | 06:52 | |
*** tshefi has quit IRC | 06:54 | |
*** mabrams has quit IRC | 07:10 | |
*** jaosorior_away is now known as jaosorior | 07:18 | |
*** mabrams has joined #openstack-glance | 07:21 | |
*** tshefi has joined #openstack-glance | 07:25 | |
*** mabrams has quit IRC | 07:28 | |
*** mingyu has quit IRC | 07:41 | |
*** mabrams has joined #openstack-glance | 07:46 | |
*** zhurong has quit IRC | 07:55 | |
*** abhishekk has joined #openstack-glance | 07:57 | |
*** mfedosin has joined #openstack-glance | 07:57 | |
*** zzzeek has quit IRC | 08:00 | |
*** zzzeek has joined #openstack-glance | 08:00 | |
*** e0ne has joined #openstack-glance | 08:01 | |
*** mingyu has joined #openstack-glance | 08:07 | |
*** aarefiev_afk is now known as aarefiev | 08:07 | |
*** tshefi has quit IRC | 08:25 | |
*** markvoelker has joined #openstack-glance | 08:26 | |
*** mingyu has quit IRC | 08:26 | |
*** mingyu has joined #openstack-glance | 08:28 | |
*** tshefi has joined #openstack-glance | 08:34 | |
*** mingyu has quit IRC | 08:40 | |
*** mingyu has joined #openstack-glance | 08:50 | |
*** aavraham has quit IRC | 08:51 | |
*** markvoelker has quit IRC | 08:59 | |
*** mingyu has quit IRC | 09:01 | |
*** zhurong has joined #openstack-glance | 09:06 | |
*** _dalgaaf has quit IRC | 09:26 | |
*** dalgaaf has joined #openstack-glance | 09:28 | |
*** mvk has joined #openstack-glance | 09:33 | |
*** e0ne has quit IRC | 09:35 | |
*** e0ne has joined #openstack-glance | 09:39 | |
*** markvoelker has joined #openstack-glance | 09:57 | |
*** tovin07_ has quit IRC | 09:59 | |
*** namnh has quit IRC | 10:01 | |
*** mingyu has joined #openstack-glance | 10:01 | |
*** mingyu has quit IRC | 10:06 | |
*** ratailor has quit IRC | 10:14 | |
*** zhurong has quit IRC | 10:20 | |
*** ratailor has joined #openstack-glance | 10:27 | |
*** markvoelker has quit IRC | 10:30 | |
*** smatzek has joined #openstack-glance | 10:30 | |
*** abhishekk has quit IRC | 10:36 | |
*** zhurong has joined #openstack-glance | 10:41 | |
*** smatzek has quit IRC | 10:42 | |
*** sdague has joined #openstack-glance | 10:49 | |
*** bkopilov has quit IRC | 10:52 | |
*** pkovar has joined #openstack-glance | 11:05 | |
*** pkovar has quit IRC | 11:15 | |
*** pkovar has joined #openstack-glance | 11:18 | |
*** udesale has quit IRC | 11:27 | |
*** markvoelker has joined #openstack-glance | 11:27 | |
*** gabor_antal has joined #openstack-glance | 11:28 | |
*** smatzek has joined #openstack-glance | 11:38 | |
*** mine0901 has quit IRC | 11:40 | |
*** chlong has joined #openstack-glance | 11:57 | |
*** ratailor has quit IRC | 11:58 | |
*** markvoelker has quit IRC | 11:59 | |
*** ratailor has joined #openstack-glance | 12:16 | |
*** openstackgerrit has quit IRC | 12:18 | |
*** zhurong has quit IRC | 12:19 | |
*** pkovar has quit IRC | 12:21 | |
*** markvoelker has joined #openstack-glance | 12:25 | |
*** bkopilov has joined #openstack-glance | 12:31 | |
*** catintheroof has joined #openstack-glance | 12:33 | |
*** jdurgin has joined #openstack-glance | 12:53 | |
*** flaper87 has quit IRC | 13:06 | |
*** lucasxu has joined #openstack-glance | 13:18 | |
*** ratailor has quit IRC | 13:25 | |
*** jaosorior has quit IRC | 13:32 | |
*** jaosorior has joined #openstack-glance | 13:32 | |
*** links has quit IRC | 13:33 | |
*** clayton has quit IRC | 13:35 | |
*** mfedosin has quit IRC | 13:36 | |
*** mfedosin has joined #openstack-glance | 13:37 | |
*** clayton has joined #openstack-glance | 13:37 | |
*** flaper87 has joined #openstack-glance | 14:14 | |
*** flaper87 has quit IRC | 14:15 | |
*** flaper87 has joined #openstack-glance | 14:15 | |
*** openstackgerrit has joined #openstack-glance | 14:17 | |
openstackgerrit | Sean McGinnis proposed openstack/glance master: Remove use of config enforce_type=True https://review.openstack.org/473622 | 14:17 |
*** mine0901 has joined #openstack-glance | 14:36 | |
*** jdurgin has quit IRC | 14:39 | |
sdague | anyone up for discussing the merrits of this patch - https://review.openstack.org/#/c/468443/ on the global_request_id work? | 14:43 |
*** jdurgin has joined #openstack-glance | 14:45 | |
jokke_ | sdague: I'm here | 14:48 |
jokke_ | abhishek_k: around? | 14:48 |
openstackgerrit | Merged openstack/glance master: Log at error when we intend to reraise the exception https://review.openstack.org/417222 | 14:48 |
openstackgerrit | Merged openstack/glance master: Trivial fix https://review.openstack.org/463488 | 14:49 |
sdague | hey jokke_, do you understand my concerns about overwriting both values? | 14:53 |
jokke_ | sdague: yes and no | 14:53 |
sdague | because if we do, it will be much harder to tell when nova / cinder do a complicated flow in glance, that things were multiple API calls being processed | 14:53 |
jokke_ | so 1st abhishek_k actually recalled specific behavior that will make it absolute mess if we don't set the req_id when we are actually sending it | 14:54 |
sdague | jokke_: ok, explain? | 14:54 |
jokke_ | so what prevents cinder and nova sending the global_req_id to glance? | 14:54 |
sdague | jokke_: I'm not sure I understand you question | 14:55 |
jokke_ | for example when we paginate the listing of images ... we did specifically decide to utilize the same req id so that all your listing calls are easily identifiable. If we now change from setting the req_id to setting global_req_id we will get following: | 14:56 |
jokke_ | 1) glance client requests first page if list and gets req id back from glance | 14:56 |
sdague | when who paginates that? | 14:57 |
jokke_ | 2) glance client requests next page and sets the req_id it received as the req id, which glance places to global_req_id and returns new req_id | 14:57 |
jokke_ | 3) glance client uses that new req_id when requesting next page that gets now set to global and again new req_id gets returned | 14:58 |
sdague | ok. So that's *very* different behavior than the rest of openstack, but that being said, what would happen after this would be that global_request_id would be there for subsequent calls | 14:58 |
jokke_ | so if we utilize it already within glance we will break the api for anyone who has been utilizing the same on their own workflow | 14:58 |
sdague | so, your concern is for applications that depend on glance returning the same request id as the thing that's sent in? | 14:59 |
sdague | instead of storing it in a local variable | 14:59 |
jokke_ | sdague: well that's the problem ... we did tell everyone else that we have req_ids and possibility to set them already in when they were rolled to rest of the openstack. And our concerns of deviation were just ignored at that time | 15:00 |
*** mabrams has quit IRC | 15:00 | |
sdague | jokke_: sure, but that ship is very much saild | 15:00 |
sdague | sailed | 15:00 |
jokke_ | sdague: yes, glance allows the client (regardless what client is used) to send req_id and returns that same req_id back | 15:01 |
sdague | so, basically we are in a situation where glance can be more like other base services, making the interactions with the rest of neutron / cinder / nova more similar | 15:01 |
sdague | or it can continue to be different and have to be coded around | 15:01 |
*** danpawlik is now known as _danpawlik | 15:02 | |
jokke_ | now we're just fine to take that global_req_id onboard and align it's behavior (apart from setting global when the "local" is sent in the request) so if we want to have global_req_id to be set, we need to send global_req_id not req_id | 15:02 |
sdague | jokke_: by using a different header than the rest of OpenStack, it is still different | 15:02 |
jokke_ | and this is the moment where we can align everyone else to that as it's not done yet | 15:03 |
jokke_ | sdague: so lets send the global for everyone | 15:03 |
sdague | it's all done | 15:03 |
jokke_ | and not fuck it up for glance | 15:03 |
sdague | litterally, everyone else is done and shipped | 15:03 |
jokke_ | well then it needs to be worked around as we were again not listened in the first place | 15:04 |
jokke_ | we've been shipping this behavior in the API for multiple releases | 15:04 |
sdague | jokke_: there have been many other behavior changes that have happened in that time | 15:07 |
sdague | I guess I don't understand the reason to need to be different here | 15:07 |
sdague | but, we seem to be at an impass | 15:07 |
sdague | there was quite a bit of ML traffic sent about this since Boston, it would have been nice to have heard objections ealier there | 15:08 |
*** tshefi has quit IRC | 15:08 | |
*** tesseract has quit IRC | 15:10 | |
jokke_ | so i raised this very concern in Boston in that room where we were discussing this behavior and said that the proposal will not work for glance, I did think I was pretty clear on the review as well ... the fact that situation in Glance is what it is and I haven't been able to follow every conversation in the ML not directly flagged to me is unfortunate but it is what it is | 15:12 |
sdague | ok, so I guess the question is whether you feel that is your pov, or the glance core team concensus | 15:12 |
*** clayton has quit IRC | 15:13 | |
sdague | because if it's the concesus, I'll drop it and move on to other projects | 15:13 |
sdague | it will be unfortunate that any request_id tooling won't work as well for glance, but we'll just have to document that fact | 15:13 |
jokke_ | as said I have no problem to accept global_req_id for glance when one is sent | 15:14 |
*** clayton has joined #openstack-glance | 15:15 | |
jokke_ | it's easy solution and not need to break the current behavior of the API just for the sake of breaking it | 15:16 |
sdague | right, that's all happening right now already | 15:16 |
jokke_ | like? | 15:16 |
sdague | it just means the API tree structure can't be reconstructed | 15:16 |
*** clayton has quit IRC | 15:43 | |
*** clayton has joined #openstack-glance | 15:48 | |
*** d0ugal has quit IRC | 15:53 | |
*** jaosorior has quit IRC | 15:57 | |
*** jaosorior has joined #openstack-glance | 15:58 | |
*** clayton has quit IRC | 15:58 | |
*** jaosorior is now known as jaosorior_away | 15:58 | |
openstackgerrit | Sean McGinnis proposed openstack/glance master: Remove deprecated use_user_token related opts https://review.openstack.org/422689 | 15:58 |
*** clayton has joined #openstack-glance | 15:59 | |
*** e0ne has quit IRC | 16:00 | |
*** gyee has joined #openstack-glance | 16:00 | |
*** groen692 has quit IRC | 16:06 | |
*** jdurgin has quit IRC | 16:13 | |
*** jaosorior_away has quit IRC | 16:23 | |
*** rcernin has quit IRC | 16:41 | |
openstackgerrit | Merged openstack/glance master: Remove duplicate key from dictionary https://review.openstack.org/470170 | 16:53 |
*** jdurgin has joined #openstack-glance | 17:09 | |
*** mingyu has joined #openstack-glance | 17:27 | |
openstackgerrit | Merged openstack/glance master: Remove use of config enforce_type=True https://review.openstack.org/473622 | 17:33 |
openstackgerrit | Sean McGinnis proposed openstack/glance master: Remove use of deprecated LOG.warn https://review.openstack.org/473922 | 17:34 |
openstackgerrit | Sean McGinnis proposed openstack/glance master: Remove use of deprecated LOG.warn https://review.openstack.org/473922 | 18:10 |
*** rcernin has joined #openstack-glance | 18:26 | |
*** nicolasbock has quit IRC | 18:51 | |
*** nicolasbock has joined #openstack-glance | 18:51 | |
*** jdurgin has quit IRC | 19:07 | |
*** NostawRm has joined #openstack-glance | 19:10 | |
*** mvk has quit IRC | 19:10 | |
*** mingyu has quit IRC | 19:24 | |
*** adisky__ has quit IRC | 19:26 | |
*** jdurgin has joined #openstack-glance | 19:27 | |
*** nicolasbock has quit IRC | 19:30 | |
*** mfedosin has quit IRC | 19:32 | |
*** ericyoung has joined #openstack-glance | 19:55 | |
ericyoung | according to the doc, using Cinder as a backend for Glance is experimental. Any idea when that will no longer be the case? | 20:14 |
smcginnis | ericyoung: I've heard of people successfully using it, but I have not tried it myself. | 20:15 |
ericyoung | smcginnis: I've set it up and the nova tempest tests seem to all pass (with ScaleIO as the backend to Cinder) | 20:16 |
ericyoung | my customers are asking as they don't like taking experimental things into production | 20:17 |
smcginnis | ericyoung: Understandable. :) | 20:17 |
smcginnis | ericyoung: Could just be the case no one has gotten around to updating the doc. | 20:17 |
smcginnis | ericyoung: But I don't really know the real state myself, so this is just conjecture. | 20:17 |
* ericyoung has his fingers crossed | 20:17 | |
*** smatzek has quit IRC | 20:48 | |
*** lucasxu has quit IRC | 20:54 | |
*** mvk has joined #openstack-glance | 21:25 | |
*** rcernin has quit IRC | 21:54 | |
*** jdurgin has quit IRC | 22:11 | |
*** catintheroof has quit IRC | 22:47 | |
*** sdague has quit IRC | 22:51 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/glance master: Updated from global requirements https://review.openstack.org/472804 | 22:53 |
openstackgerrit | Merged openstack/glance_store master: Initialize privsep root_helper command https://review.openstack.org/458201 | 23:21 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!