| *** bkopilov has joined #openstack-cinder | 00:00 | |
| *** jessegler has quit IRC | 00:04 | |
| *** wxy- has joined #openstack-cinder | 00:07 | |
| *** Apoorva_ has quit IRC | 00:07 | |
| *** Apoorva has joined #openstack-cinder | 00:08 | |
| *** wxy- has quit IRC | 00:11 | |
| *** Apoorva has quit IRC | 00:12 | |
| *** gcb has quit IRC | 00:21 | |
| *** stakeda has joined #openstack-cinder | 00:27 | |
| *** edmondsw has joined #openstack-cinder | 00:49 | |
| *** edmondsw has quit IRC | 00:53 | |
| *** wxy- has joined #openstack-cinder | 00:54 | |
| *** bkopilov has quit IRC | 00:54 | |
| openstackgerrit | zheng yin proposed openstack/cinder master: Ceph backup: eventlet.green.subprocess replace subprocess https://review.openstack.org/466607 | 00:56 |
|---|---|---|
| *** bkopilov has joined #openstack-cinder | 00:57 | |
| *** wxy- has quit IRC | 00:58 | |
| *** zzzeek_ has quit IRC | 01:09 | |
| *** zzzeek_ has joined #openstack-cinder | 01:11 | |
| *** wanghao_ has joined #openstack-cinder | 01:11 | |
| *** wanghao has quit IRC | 01:15 | |
| *** bkopilov has quit IRC | 01:16 | |
| *** bkopilov has joined #openstack-cinder | 01:16 | |
| *** sdague has quit IRC | 01:17 | |
| *** Apoorva has joined #openstack-cinder | 01:18 | |
| *** wanghao has joined #openstack-cinder | 01:20 | |
| *** wanghao_ has quit IRC | 01:23 | |
| openstackgerrit | liuyamin proposed openstack/cinder master: Use QoSConsumerValues enum field https://review.openstack.org/479674 | 01:26 |
| *** wanghao_ has joined #openstack-cinder | 01:28 | |
| *** wanghao has quit IRC | 01:31 | |
| *** zzzeek_ has quit IRC | 01:33 | |
| *** zzzeek_ has joined #openstack-cinder | 01:34 | |
| *** zzzeek_ has quit IRC | 01:36 | |
| *** Shunli has joined #openstack-cinder | 01:41 | |
| *** zzzeek_ has joined #openstack-cinder | 01:47 | |
| *** Apoorva has quit IRC | 01:49 | |
| *** gkadam has joined #openstack-cinder | 01:49 | |
| *** Apoorva has joined #openstack-cinder | 01:49 | |
| *** abhitechie has joined #openstack-cinder | 01:52 | |
| *** Apoorva has quit IRC | 01:54 | |
| *** bkopilov has quit IRC | 01:59 | |
| *** zhaochao has joined #openstack-cinder | 02:00 | |
| *** gcb has joined #openstack-cinder | 02:03 | |
| *** zhurong has joined #openstack-cinder | 02:03 | |
| openstackgerrit | TommyLike proposed openstack/cinder master: Enable backend specific async migration https://review.openstack.org/334923 | 02:15 |
| jgriffith | mriedem sorry, I didn't see your follow up | 02:21 |
| jgriffith | mriedem so yes, my point was that as it stands we should probably add the reserved status to the expected list | 02:21 |
| jgriffith | otherwise there are windows of failure | 02:21 |
| jgriffith | mriedem if we're not comfortable with doing that that's fine as well, but we need to comment the code and make sure it's obvious in the message | 02:22 |
| jgriffith | I would hope that people wouldn't be extending without some thought, but I guess you never know | 02:22 |
| jgriffith | The existing exception mesg is probably good enough, but would be helpful to have some more detail. | 02:23 |
| jgriffith | I'll update the patch, but I'm curious what your thoughts are around allowing extend on reserved? | 02:23 |
| jgriffith | I'm not sure there's a real risk to it? | 02:23 |
| jgriffith | I'll remove my -1 and +2 when you and Mathieu let me know which behavior we want and we document it. | 02:27 |
| jgriffith | mgagne ^^ | 02:34 |
| *** edmondsw has joined #openstack-cinder | 02:37 | |
| *** edmondsw has quit IRC | 02:42 | |
| *** abhitechie has quit IRC | 02:57 | |
| *** wanghao_ has quit IRC | 02:58 | |
| openstackgerrit | Liu Qing proposed openstack/cinder master: Fix SSHPool current_size not work correctly https://review.openstack.org/468634 | 02:58 |
| *** wanghao has joined #openstack-cinder | 02:59 | |
| openstackgerrit | Bin Zhou proposed openstack/cinder master: Support utf-8 decode https://review.openstack.org/482389 | 03:03 |
| openstackgerrit | Bin Zhou proposed openstack/cinder master: Support utf-8 decode https://review.openstack.org/482389 | 03:07 |
| openstackgerrit | Bin Zhou proposed openstack/cinder master: Support utf-8 decode https://review.openstack.org/482389 | 03:09 |
| *** bkopilov has joined #openstack-cinder | 03:18 | |
| *** abhitechie has joined #openstack-cinder | 03:21 | |
| *** diablo_rojo has quit IRC | 03:25 | |
| *** wanghao has quit IRC | 03:27 | |
| *** wanghao has joined #openstack-cinder | 03:28 | |
| *** abhitechie has quit IRC | 03:35 | |
| *** abhitechie has joined #openstack-cinder | 03:36 | |
| *** links has joined #openstack-cinder | 03:41 | |
| *** abhitechie has quit IRC | 03:46 | |
| *** jmlowe has joined #openstack-cinder | 03:58 | |
| *** psachin has joined #openstack-cinder | 04:00 | |
| *** dave-mccowan has quit IRC | 04:00 | |
| *** isaacb has joined #openstack-cinder | 04:08 | |
| *** isaacb has quit IRC | 04:10 | |
| *** abhitechie has joined #openstack-cinder | 04:15 | |
| openstackgerrit | kongxiangyun proposed openstack/cinder master: add exception check for lvm_type https://review.openstack.org/481454 | 04:18 |
| *** edmondsw has joined #openstack-cinder | 04:25 | |
| *** edmondsw has quit IRC | 04:30 | |
| *** abhitechie has quit IRC | 04:30 | |
| *** abhitechie has joined #openstack-cinder | 04:30 | |
| *** arnewiebalck_ has joined #openstack-cinder | 04:33 | |
| *** Julien-zte has quit IRC | 04:41 | |
| *** Julien-zte has joined #openstack-cinder | 04:42 | |
| *** markstur has quit IRC | 04:55 | |
| openstackgerrit | iswarya vakati proposed openstack/cinder master: Stop using deprecated 'message' attribute in Exception https://review.openstack.org/482125 | 04:57 |
| *** abhitechie has quit IRC | 04:57 | |
| *** abhinavtechie has joined #openstack-cinder | 04:57 | |
| *** abhitechie has joined #openstack-cinder | 05:28 | |
| *** abhinavtechie has quit IRC | 05:28 | |
| *** abhitechie has quit IRC | 05:32 | |
| *** abhitechie has joined #openstack-cinder | 05:34 | |
| *** Shunli has quit IRC | 05:39 | |
| *** Shunli has joined #openstack-cinder | 05:40 | |
| *** rcernin has joined #openstack-cinder | 05:41 | |
| *** lpetrut has joined #openstack-cinder | 05:41 | |
| *** arnewiebalck_ has quit IRC | 05:49 | |
| *** nicolasbock has joined #openstack-cinder | 05:55 | |
| *** gkadam has quit IRC | 05:56 | |
| *** gkadam has joined #openstack-cinder | 05:58 | |
| *** pcaruana has joined #openstack-cinder | 06:04 | |
| *** ChubYann has quit IRC | 06:07 | |
| *** abhitechie has quit IRC | 06:09 | |
| *** zhonghua has quit IRC | 06:11 | |
| *** zhonghua has joined #openstack-cinder | 06:12 | |
| *** edmondsw has joined #openstack-cinder | 06:14 | |
| *** abhitechie has joined #openstack-cinder | 06:15 | |
| *** gkadam has quit IRC | 06:17 | |
| *** gkadam has joined #openstack-cinder | 06:17 | |
| *** edmondsw has quit IRC | 06:18 | |
| *** gkadam has quit IRC | 06:18 | |
| *** gkadam has joined #openstack-cinder | 06:19 | |
| *** gkadam has quit IRC | 06:21 | |
| *** jmlowe has quit IRC | 06:25 | |
| *** laurent\ has quit IRC | 06:35 | |
| *** laurent\ has joined #openstack-cinder | 06:36 | |
| *** abhitechie has quit IRC | 06:36 | |
| *** abhitechie has joined #openstack-cinder | 06:36 | |
| *** lpetrut has quit IRC | 06:39 | |
| *** zhurong has quit IRC | 06:40 | |
| *** gkadam has joined #openstack-cinder | 06:44 | |
| *** zhurong has joined #openstack-cinder | 06:44 | |
| *** arnewiebalck_ has joined #openstack-cinder | 06:45 | |
| *** belmoreira has joined #openstack-cinder | 06:45 | |
| *** gkadam has quit IRC | 06:46 | |
| *** lhx__ has joined #openstack-cinder | 06:46 | |
| *** arnewiebalck_ has quit IRC | 06:46 | |
| *** Poornima has joined #openstack-cinder | 06:47 | |
| *** arnewiebalck_ has joined #openstack-cinder | 06:49 | |
| *** Poornima has quit IRC | 06:49 | |
| *** lhx__ has quit IRC | 06:50 | |
| *** lhx__ has joined #openstack-cinder | 06:50 | |
| *** tesseract has joined #openstack-cinder | 07:02 | |
| *** lkuchlan has joined #openstack-cinder | 07:09 | |
| *** zhurong has quit IRC | 07:30 | |
| *** mriedem has quit IRC | 07:30 | |
| *** openstackgerrit has quit IRC | 07:33 | |
| *** openstackgerrit has joined #openstack-cinder | 07:42 | |
| openstackgerrit | kongxiangyun proposed openstack/cinder master: add exception check for lvm_type https://review.openstack.org/481454 | 07:42 |
| *** aarefiev_pto is now known as aarefiev | 07:44 | |
| *** markvoelker_ has quit IRC | 07:47 | |
| *** zzzeek has quit IRC | 08:00 | |
| *** zzzeek has joined #openstack-cinder | 08:00 | |
| *** openstackgerrit has quit IRC | 08:03 | |
| *** abhitechie has quit IRC | 08:12 | |
| *** openstackgerrit has joined #openstack-cinder | 08:17 | |
| openstackgerrit | Arnon Yaari proposed openstack/cinder master: INFINIDAT: Add QoS support https://review.openstack.org/482132 | 08:17 |
| *** Nishant_ has joined #openstack-cinder | 08:18 | |
| *** e0ne has joined #openstack-cinder | 08:19 | |
| *** abhitechie has joined #openstack-cinder | 08:23 | |
| *** isaacb has joined #openstack-cinder | 08:23 | |
| *** gkadam has joined #openstack-cinder | 08:25 | |
| *** mvk has quit IRC | 08:26 | |
| *** gkadam is now known as gkadam-afk | 08:26 | |
| *** gkadam-afk has quit IRC | 08:26 | |
| *** zhurong has joined #openstack-cinder | 08:27 | |
| *** mvk has joined #openstack-cinder | 08:55 | |
| openstackgerrit | Helen Walsh proposed openstack/cinder master: VMAX driver - seamless upgrade from SMI-S to REST https://review.openstack.org/482138 | 09:00 |
| *** lpetrut has joined #openstack-cinder | 09:04 | |
| *** pgadiya has joined #openstack-cinder | 09:13 | |
| *** arnewiebalck_ has quit IRC | 09:15 | |
| *** zhurong has quit IRC | 09:27 | |
| *** Shunli has quit IRC | 09:29 | |
| *** e0ne has quit IRC | 09:31 | |
| *** e0ne has joined #openstack-cinder | 09:34 | |
| *** mvk has quit IRC | 09:37 | |
| *** links has quit IRC | 09:37 | |
| *** links has joined #openstack-cinder | 09:39 | |
| *** links has quit IRC | 09:44 | |
| *** dalgaaf has quit IRC | 09:47 | |
| *** dalgaaf has joined #openstack-cinder | 09:47 | |
| *** markvoelker has joined #openstack-cinder | 09:48 | |
| *** links has joined #openstack-cinder | 09:49 | |
| *** edmondsw has joined #openstack-cinder | 09:50 | |
| *** mvk has joined #openstack-cinder | 09:51 | |
| *** edmondsw has quit IRC | 09:54 | |
| *** kongwei has joined #openstack-cinder | 09:58 | |
| *** kongwei has quit IRC | 10:00 | |
| openstackgerrit | TommyLike proposed openstack/cinder master: [WIP] Implement policies in code, reviews are welcome https://review.openstack.org/482492 | 10:03 |
| *** wxy- has joined #openstack-cinder | 10:04 | |
| *** sdague has joined #openstack-cinder | 10:05 | |
| *** wanghao has quit IRC | 10:12 | |
| *** wanghao has joined #openstack-cinder | 10:13 | |
| *** gcb has quit IRC | 10:13 | |
| *** links has quit IRC | 10:13 | |
| *** wanghao has quit IRC | 10:17 | |
| *** wanghao has joined #openstack-cinder | 10:17 | |
| *** markvoelker has quit IRC | 10:22 | |
| *** wxy-_ has joined #openstack-cinder | 10:29 | |
| *** wxy- has quit IRC | 10:29 | |
| *** gouthamr has quit IRC | 10:33 | |
| openstackgerrit | Yu Shengzuo proposed openstack/cinder master: Fix no feature to enable 'exclusive-lock' of image in ceph https://review.openstack.org/481472 | 10:38 |
| *** gouthamr has joined #openstack-cinder | 10:41 | |
| kaisers | Hey! Anyone with exp in encrypted volumes and a spare moment might help with taking a look at the questions in this bug: https://bugs.launchpad.net/tempest/+bug/1702843 , please | 10:58 |
| openstack | Launchpad bug 1702843 in tempest "test_boot_server_from_encrypted_volume_luks has insufficient skip decorator" [Undecided,New] | 10:58 |
| *** gkadam has joined #openstack-cinder | 11:00 | |
| *** dave-mccowan has joined #openstack-cinder | 11:10 | |
| *** wxy- has joined #openstack-cinder | 11:10 | |
| *** bkopilov has quit IRC | 11:11 | |
| *** Julien-zte has quit IRC | 11:12 | |
| *** wxy- has quit IRC | 11:13 | |
| *** Julien-zte has joined #openstack-cinder | 11:13 | |
| *** wxy-_ has quit IRC | 11:14 | |
| *** markvoelker has joined #openstack-cinder | 11:20 | |
| *** gkadam has quit IRC | 11:24 | |
| *** edmondsw has joined #openstack-cinder | 11:38 | |
| *** wanghao has quit IRC | 11:38 | |
| *** abishop has joined #openstack-cinder | 11:40 | |
| *** edmondsw has quit IRC | 11:42 | |
| *** links has joined #openstack-cinder | 11:46 | |
| openstackgerrit | Vivek Soni proposed openstack/cinder master: 3PAR: Add volume to consistency/generic group https://review.openstack.org/480255 | 11:47 |
| *** markvoelker has quit IRC | 11:53 | |
| *** links has quit IRC | 11:59 | |
| *** gkadam has joined #openstack-cinder | 12:05 | |
| *** Nishant_ has quit IRC | 12:07 | |
| *** felipemonteiro_ has joined #openstack-cinder | 12:08 | |
| *** felipemonteiro__ has joined #openstack-cinder | 12:09 | |
| *** dustins has joined #openstack-cinder | 12:09 | |
| *** felipemonteiro_ has quit IRC | 12:12 | |
| *** liverpooler has joined #openstack-cinder | 12:18 | |
| *** markvoelker has joined #openstack-cinder | 12:19 | |
| *** MVenesio has joined #openstack-cinder | 12:32 | |
| *** bkopilov has joined #openstack-cinder | 12:33 | |
| *** Poornima has joined #openstack-cinder | 12:35 | |
| *** abhitechie has quit IRC | 12:38 | |
| *** mlakat has quit IRC | 12:46 | |
| *** edmondsw has joined #openstack-cinder | 12:46 | |
| *** pgadiya has quit IRC | 12:47 | |
| *** openstackgerrit has quit IRC | 12:47 | |
| *** gkadam_ has joined #openstack-cinder | 12:51 | |
| *** lhx__ has quit IRC | 12:52 | |
| *** gkadam has quit IRC | 12:54 | |
| *** chhavi has joined #openstack-cinder | 13:03 | |
| *** gkadam__ has joined #openstack-cinder | 13:04 | |
| arnewiebalck | We have just upgraded our Cinder deployment here at CERN to Ocata. | 13:04 |
| arnewiebalck | The c-vol for one of our Ceph pools did not want to start. | 13:05 |
| *** openstackgerrit has joined #openstack-cinder | 13:06 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/python-cinderclient master: Updated from global requirements https://review.openstack.org/478071 | 13:06 |
| *** gkadam_ has quit IRC | 13:06 | |
| *** lkuchlan has quit IRC | 13:07 | |
| arnewiebalck | The problem was the gathering of the provisioning stats. | 13:07 |
| *** gkadam__ has quit IRC | 13:11 | |
| arnewiebalck | From what I see, the code loops over all volumes in a periodic task every 60 secs. | 13:11 |
| arnewiebalck | With more than 4’000 volumes in the pool, this takes too long and it basically enters an infinite loop querying all volumes. | 13:12 |
| arnewiebalck | We have now taken out this line: https://github.com/openstack/cinder/blob/master/cinder/volume/drivers/rbd.py#L415 | 13:15 |
| Dinesh_Bhor | arnewiebalck: I think this is related https://bugs.launchpad.net/cinder/+bug/1681374 | 13:15 |
| openstack | Launchpad bug 1681374 in Cinder "cinder-volume can't be inited with large number of volume" [Medium,Confirmed] | 13:15 |
| arnewiebalck | Dinesh_Bhor: I think this is different. | 13:16 |
| arnewiebalck | Dinesh_Bhor: Thanks for the pointer, btw ;) | 13:16 |
| Dinesh_Bhor | arnewiebalck: ok, hope someone will be answer | 13:18 |
| arnewiebalck | arnewiebalck: The loop is also triggered on deletion, it seems, not only on startup. | 13:18 |
| openstackgerrit | Rawan Herzallah proposed openstack/cinder master: Adding NVMEoF target for LVM driver https://review.openstack.org/482562 | 13:21 |
| smcginnis | Swanson: ^ | 13:22 |
| arnewiebalck | In addition to the fact this does not scale, I struggle to understand the underlying idea. | 13:24 |
| arnewiebalck | From what I see, Cinder is asking Ceph for the allocated sizes of all volumes. Cinder knows this already, no? | 13:24 |
| smcginnis | arnewiebalck: I wonder if that's the actual space used, not the logical space configured. | 13:25 |
| arnewiebalck | e0ne: ^ | 13:26 |
| smcginnis | arnewiebalck: It was added as part of the thin provisioning work: https://github.com/openstack/cinder/commit/d4fd5660736a1363a4e78480b116532c71b5ce49 | 13:26 |
| arnewiebalck | smcginnis: not it’s not | 13:26 |
| arnewiebalck | smcginnis: I checked experimentally (and had our Ceph expert check the code) | 13:26 |
| arnewiebalck | smcginnis: I’m certainly missing sth. | 13:26 |
| openstackgerrit | Chaynika Saikia proposed openstack/python-cinderclient master: Add cinder create --poll https://review.openstack.org/475919 | 13:26 |
| e0ne | arnewiebalck: what issue are you talking about? | 13:27 |
| smcginnis | arnewiebalck: Quite possible the logic may be wrong. | 13:27 |
| *** josecastroleon has joined #openstack-cinder | 13:27 | |
| arnewiebalck | e0ne: the one I described ^ | 13:27 |
| *** lhx_ has joined #openstack-cinder | 13:29 | |
| e0ne | interesting | 13:30 |
| *** gcb has joined #openstack-cinder | 13:32 | |
| *** Poornima has quit IRC | 13:34 | |
| *** wxy- has joined #openstack-cinder | 13:34 | |
| arnewiebalck | smcginnis: my “no, it’s not” referred to the actual/allocated space; you’re correct about the commit, we’ve found this as well | 13:38 |
| smcginnis | arnewiebalck: Hah, that's how I took it. ;) | 13:39 |
| *** wxy- has quit IRC | 13:39 | |
| e0ne | arnewiebalck: here is PoC how to make it faster | 13:39 |
| arnewiebalck | smcginnis: In the corresponding review, there were dicusssions about actual vs allocated space, but I didn’t get the idea in the end. | 13:40 |
| e0ne | arnewiebalck, smcginnis: sorry, forgot to paste the link: https://review.openstack.org/#/c/476079/ | 13:46 |
| arnewiebalck | e0ne: :-D | 13:48 |
| arnewiebalck | e0ne: this also changes from allocated space to used space? | 13:48 |
| e0ne | arnewiebalck: yes | 13:48 |
| *** eharney has joined #openstack-cinder | 13:48 | |
| arnewiebalck | e0ne: so, the current logic is wrong? | 13:49 |
| e0ne | but jbernard has some valid concerns on it | 13:49 |
| e0ne | arnewiebalck: I'm not sure, TBH | 13:49 |
| *** abhitechie has joined #openstack-cinder | 13:49 | |
| arnewiebalck | e0ne: yes, his point 1) is what I just asked | 13:50 |
| e0ne | arnewiebalck: we can discuss it tomorrow at the meeting | 13:50 |
| arnewiebalck | e0ne: sure | 13:51 |
| *** mlakat has joined #openstack-cinder | 13:52 | |
| arnewiebalck | e0ne: our local Ceph expert promised to check if there is a way to address Jon’s point 2) | 13:53 |
| e0ne | arnewiebalck: I didn't find any API in current ceph/rbd/rados python bindings | 13:54 |
| e0ne | arnewiebalck: I hope, you have a better luck with it | 13:54 |
| arnewiebalck | e0ne: he thinks “no”, but wanted to double check, let’s see :) | 13:55 |
| e0ne | :) | 13:55 |
| arnewiebalck | smcginnis: ok with you if we put this on the agenda for tomorrow? | 13:57 |
| *** jistr is now known as jistr|call | 13:58 | |
| smcginnis | arnewiebalck: Yes, absolutely. | 14:02 |
| arnewiebalck | smcginnis: done | 14:02 |
| smcginnis | Great | 14:03 |
| _pewp_ | hemna (ه’́⌣’̀ه )／ | 14:03 |
| smcginnis | hemna: guten morgen | 14:13 |
| hemna | :) | 14:13 |
| hemna | morning | 14:14 |
| hemna | getting up to speed on the new gig | 14:14 |
| smcginnis | hemna: I had been wondering how that was going. Glad to see the news. | 14:14 |
| hemna | It's a relief | 14:14 |
| smcginnis | hemna: I bet | 14:15 |
| eharney | hemna: glad to hear! | 14:15 |
| e0ne | hemna: evening:) | 14:15 |
| *** dfisher has joined #openstack-cinder | 14:16 | |
| hemna | at least I'll still be doing OpenStack for sure now too | 14:16 |
| *** zzzeek_ has quit IRC | 14:18 | |
| openstackgerrit | Helen Walsh proposed openstack/cinder master: VMAX driver - Replication, replacing SMI-S with REST https://review.openstack.org/466027 | 14:18 |
| ildikov | hemna: awesome news! :) | 14:19 |
| hemna | thank you :) | 14:20 |
| scottda | Congrats on your escape hemna :) | 14:20 |
| *** zzzeek_ has joined #openstack-cinder | 14:21 | |
| hemna | :) thanks scottda | 14:21 |
| smcginnis | We'll have to all drive up and kidnap scottda and bring him to Rosa Mexicana for some Manberries. | 14:22 |
| hemna | ooh yah | 14:22 |
| scottda | Ha. Where's Rosa Mexicana? Denver? | 14:23 |
| scottda | For the PTG: http://www.richardsandoval.com/tamayo/ | 14:23 |
| smcginnis | scottda: Isn't that the place in Ft Collins we went to where kmartin got the drink? | 14:24 |
| scottda | No, that was The Rio. There is one in downtown Denver BTW | 14:25 |
| dfisher | The Rio is horrible. | 14:25 |
| openstackgerrit | Helen Walsh proposed openstack/cinder master: VMAX driver - Live Migration, replacing SMI-S with REST https://review.openstack.org/472279 | 14:25 |
| smcginnis | Rio, rosa, whatever. :) | 14:25 |
| smcginnis | dfisher: Hah! | 14:25 |
| dfisher | margaritas hit like trucks | 14:25 |
| dfisher | but that's about all | 14:25 |
| dfisher | cajun food in Denver: Bayou Bob's. *amazing* | 14:25 |
| dfisher | http://www.bayoubobs.com/ — slow internet connection but the food is outstanding. | 14:26 |
| scottda | smcginnis: What is up with https://review.openstack.org/#/c/471792/? | 14:26 |
| scottda | Does it really take 48+ hours to get through the merge queue? | 14:27 |
| smcginnis | scottda: One of the dependent patches hasn't merged yet. | 14:27 |
| scottda | Or did Jenkins have to reboot or something, and it got lost? | 14:27 |
| scottda | ahh......silly me. | 14:27 |
| *** xyang1 has joined #openstack-cinder | 14:27 | |
| smcginnis | scottda: And I've got a -1 on it from someone on the tempest team because I decided not to rewrite half of tempest to fix this specific issue. | 14:27 |
| scottda | smcginnis: Yeah, I think that's called "The Tempest Blues". | 14:28 |
| *** xyang has joined #openstack-cinder | 14:28 | |
| openstackgerrit | Helen Walsh proposed openstack/cinder master: VMAX - Live Migration, replacing SMI-S with REST https://review.openstack.org/472279 | 14:28 |
| smcginnis | scottda: Not really sure at this point if I should start down that path of redoing everything, or just wait to see if another tempest core is willing to let it pass. | 14:28 |
| jgriffith | dfisher you don't go to Bayou Bob's to surf the net or work, you go for the cajun goodness, my god man, the CAJUN GOODNESS! | 14:28 |
| dfisher | amen. | 14:28 |
| smcginnis | Sounds like we have something on our evening agenda already. :) | 14:29 |
| dfisher | tryin' to help whoever's an out-of-towner to not go to the Rio. | 14:29 |
| dfisher | seriously, just go chug cuervo instead. | 14:29 |
| jgriffith | ha! | 14:29 |
| smcginnis | dfisher: Don't put it past this group. | 14:29 |
| jgriffith | smcginnis +1 | 14:29 |
| dfisher | you all work on *cinder* | 14:29 |
| dfisher | i'd be stunned if you weren't hammered right now. | 14:30 |
| smcginnis | Haha | 14:30 |
| dfisher | at least it's not Neutron ... | 14:30 |
| smcginnis | That's the truth. | 14:30 |
| smcginnis | jgriffith: Hey, wrt reserved state in that patch... | 14:31 |
| smcginnis | jgriffith: If it's reserved, that's some kind of indication that there's some work in progress, right. So if we allow extend with reserve, does that open up a race window where something starts with the old size, it gets extended mid-process, and it completes with stale information. | 14:32 |
| smcginnis | jgriffith: Not sure if that's even a possibility or something to be concerned about. | 14:32 |
| ildikov | smcginnis: in the current Nova implementation, which is still under review, you have the volume reserved in the regular attach scenarios, which can open the race window you refer to | 14:35 |
| *** gheorghe_ has quit IRC | 14:35 | |
| ildikov | smcginnis: I have no idea what happens if we do attach and extend at the same time, and I'm not sure I would want to try, but that's just my taste :) | 14:35 |
| smcginnis | ildikov: That's what I was thinking of. But will it cause problems if the volume ends up being bigger by the time it is all done than when it started off? That part I'm not sure of. | 14:35 |
| ildikov | smcginnis: the other scenario when you'll find the volume in reserved state is shelved offloaded | 14:36 |
| smcginnis | ildikov: Yep. | 14:36 |
| smcginnis | Maybe it's not a big deal, but probably something we should think about. | 14:36 |
| ildikov | smcginnis: in case of swap_volume that can be nasty I think | 14:36 |
| ildikov | as that operation specifically checks the volume size to do that operation | 14:37 |
| *** rcernin has quit IRC | 14:38 | |
| ildikov | smcginnis: I don't know what happens if a shelved offloaded server is put back to operation and gets back a bigger volume, that might not be an issue, I don't really know that part | 14:38 |
| ildikov | smcginnis: in case of swap I would think the volume has a state change in the danger zone, but when it comes to race conditions, I'm not sure whether or not I would depend on that | 14:39 |
| ildikov | smcginnis: my 2 cents :) | 14:39 |
| smcginnis | ildikov: Probably an extremely unlikely case anyway. | 14:39 |
| ildikov | smcginnis: and that's what guarantees you that it'll happen in production... :) | 14:40 |
| ildikov | smcginnis: this or something very similar | 14:40 |
| smcginnis | True! | 14:40 |
| *** rcernin has joined #openstack-cinder | 14:40 | |
| ildikov | smcginnis: does the reserved state have to be included now or it can be a decision later as well to add? | 14:42 |
| dfisher | quick question on image-to-volume: If my Glance image is gzip'd, does qemu-img decompress it when streaming the image to the volume? | 14:43 |
| openstackgerrit | Helen Walsh proposed openstack/cinder master: VMAX driver - implement Generic Volume Group feature https://review.openstack.org/474546 | 14:43 |
| smcginnis | dfisher: Yes, it has to be downloaded locally, then uncompressed. | 14:45 |
| * dfisher nods | 14:45 | |
| smcginnis | ildikov: I'd actually rather wait, but the problem with that is I think we would need to bump the microversion if we changed it afterwards. | 14:45 |
| jgriffith | smcginnis ildikov just add it if you ask me; I just wanted to get mgagne 's input first | 14:46 |
| smcginnis | jgriffith: Any thoughts on race conditions there? | 14:47 |
| jgriffith | meh... | 14:47 |
| jgriffith | we could always spend some time and "look" at the nova code and make sure we do a fresh volume-get when finalizing the attach | 14:47 |
| jgriffith | which I believe we do now | 14:47 |
| *** openstackgerrit has quit IRC | 14:48 | |
| jgriffith | I'm not as concerned about races, I raised it in the review as a potential concern, but I'm not too excited about it | 14:48 |
| smcginnis | jgriffith: It does seem like (assuming it could be a problem) it would be a very small window and very unlikely to ever happen. | 14:48 |
| jgriffith | personally I don't care which way it goes, just want it to be excplicit and documented one way or the other | 14:49 |
| smcginnis | But like ildikov said, that almost guarantees it would. | 14:49 |
| smcginnis | jgriffith: +1 | 14:49 |
| jgriffith | smcginnis ildikov haha, yeah, well there is that! | 14:49 |
| jgriffith | ahhh | 14:50 |
| jgriffith | mriedemann responded to my comments and prefers NOT to add it, so that's fine by me | 14:50 |
| smcginnis | jgriffith: Doesn't seem like we lose any major usability by not allowing it. | 14:51 |
| jgriffith | Yeah, honestly it's a corner case, not supported, get over it and move along :) | 14:52 |
| jgriffith | that's me responding to a bug in LP :) | 14:52 |
| jgriffith | finsh the attach, then extend it | 14:52 |
| jgriffith | whatever | 14:52 |
| jgriffith | it's so easy to deal with it's almost a non-issue IMO. I just want to make sure we state clearly in the api doc and the cli help that you're not allowed to do it | 14:53 |
| ildikov | smcginnis: jgriffith: I just felt someone needs to represent the worrying side | 14:55 |
| ildikov | smcginnis: jgriffith: it feels better to have the volume in a stable state, which I would not consider 'reserved' to be | 14:55 |
| ildikov | smcginnis: jgriffith: but we don't code based on feelings so you might as well ignore me :) | 14:56 |
| *** jmlowe has joined #openstack-cinder | 14:56 | |
| jgriffith | ildikov yeah, and the name "reserved" that implies you shouldn't be screwing with it anyway; somebody "reserve" the right to do that later :) | 14:56 |
| ildikov | jgriffith: exactly :) | 14:57 |
| *** isaacb has quit IRC | 14:59 | |
| *** Julien-zte has quit IRC | 15:03 | |
| *** markstur has joined #openstack-cinder | 15:06 | |
| *** diablo_rojo has joined #openstack-cinder | 15:08 | |
| *** zzzeek_ has quit IRC | 15:12 | |
| lpetrut | hi, I have two small bug fixes targeting the Windows drivers. Could anyone please take a look? Our CI is passing. https://review.openstack.org/#/c/464659/ https://review.openstack.org/#/c/462902/ | 15:12 |
| *** abhitechie has quit IRC | 15:16 | |
| *** abhitechie has joined #openstack-cinder | 15:17 | |
| *** psachin has quit IRC | 15:18 | |
| lpetrut | that was fast, thanks :) | 15:18 |
| smcginnis | lpetrut: :) | 15:18 |
| jungleboyj | Ask and ye shall receive. | 15:18 |
| *** rcernin has quit IRC | 15:18 | |
| smcginnis | lpetrut: I had actually reviewed the update on the first one but I think at the time I was just waiting for the Windows CI results. Never got back to it. | 15:18 |
| *** jistr|call is now known as jistr | 15:22 | |
| *** isaacb has joined #openstack-cinder | 15:23 | |
| jungleboyj | smcginnis: Morning sir. | 15:24 |
| xyang | jungleboyj: I have a question regarding the doc | 15:25 |
| smcginnis | jungleboyj: Hey | 15:25 |
| xyang | jungleboyj: what happened to config reference? | 15:25 |
| jungleboyj | xyang: Sure. | 15:25 |
| jungleboyj | xyang: Is it gone? | 15:25 |
| xyang | jungleboyj: yes | 15:26 |
| xyang | jungleboyj: no longer in openstack-manuals | 15:26 |
| xyang | https://github.com/openstack/openstack-manuals/tree/master/doc | 15:26 |
| xyang | but I don't see it under cinder repo either | 15:26 |
| jungleboyj | xyang: Hmmm, ok, let me look here. | 15:27 |
| *** abhinavtechie has joined #openstack-cinder | 15:27 | |
| xyang | jungleboyj: ok, thanks | 15:27 |
| jungleboyj | smcginnis: I have a related question on docs give me a minute to figure out what happened to the config reference. | 15:27 |
| *** catintheroof has joined #openstack-cinder | 15:28 | |
| smcginnis | jungleboyj: Doesn't that need to get merged and the redirect put in place? | 15:28 |
| smcginnis | I thought most of the docs.o.o links were broken for now until everything gets moved over. | 15:28 |
| jungleboyj | smcginnis: Yuck, the config-reference wasn't in the list of things we needed to move over. | 15:29 |
| smcginnis | Ruh roh | 15:29 |
| jungleboyj | smcginnis: But it was removed. | 15:29 |
| jungleboyj | Let me check with Doug but it looks like I will need to migrate that as well. | 15:30 |
| jungleboyj | xyang: ^^^ | 15:30 |
| *** zzzeek_ has joined #openstack-cinder | 15:30 | |
| *** abhitechie has quit IRC | 15:30 | |
| *** e0ne has quit IRC | 15:34 | |
| xyang | jungleboyj: thanks:) | 15:35 |
| *** belmoreira has quit IRC | 15:38 | |
| jungleboyj | xyang: It is now supposed to be autogenerated. Doug pointed me to an example. I will try to do the same for Cinder here today. | 15:40 |
| xyang | jungleboyj: Great. thanks | 15:42 |
| jungleboyj | xyang: Thanks for catching it. Let me know as you find other things. | 15:43 |
| *** abhinavtechie has quit IRC | 15:44 | |
| *** liverpooler has quit IRC | 15:44 | |
| *** lpetrut has quit IRC | 15:56 | |
| *** aarefiev is now known as aarefiev_afk | 15:56 | |
| *** jmlowe has quit IRC | 15:57 | |
| *** abishop_ has joined #openstack-cinder | 15:58 | |
| *** abishop has quit IRC | 15:59 | |
| *** eharney has quit IRC | 16:00 | |
| *** jmlowe has joined #openstack-cinder | 16:00 | |
| *** markstur has quit IRC | 16:01 | |
| *** markstur has joined #openstack-cinder | 16:02 | |
| *** Apoorva has joined #openstack-cinder | 16:08 | |
| *** openstackgerrit has joined #openstack-cinder | 16:10 | |
| openstackgerrit | Yu Shengzuo proposed openstack/cinder master: Fix no feature to enable 'exclusive-lock' of image in ceph https://review.openstack.org/481472 | 16:10 |
| openstackgerrit | Apoorva Deshpande proposed openstack/cinder master: Tintri: Fix config option access https://review.openstack.org/482278 | 16:11 |
| *** mriedem has joined #openstack-cinder | 16:11 | |
| *** isaacb has quit IRC | 16:16 | |
| *** edmondsw has quit IRC | 16:17 | |
| *** lhx_ has quit IRC | 16:17 | |
| *** edmondsw_ has joined #openstack-cinder | 16:19 | |
| openstackgerrit | Helen Walsh proposed openstack/cinder master: VMAX driver - implement Generic Volume Group feature https://review.openstack.org/474546 | 16:21 |
| *** jmlowe has quit IRC | 16:22 | |
| *** edmondsw_ has quit IRC | 16:23 | |
| *** liverpooler has joined #openstack-cinder | 16:24 | |
| openstackgerrit | Helen Walsh proposed openstack/cinder master: VMAX driver - bootable volume errors https://review.openstack.org/472674 | 16:24 |
| openstackgerrit | Helen Walsh proposed openstack/cinder master: VMAX driver - None connector object in a terminate_connection in Pike https://review.openstack.org/482130 | 16:27 |
| * jungleboyj sighs | 16:27 | |
| jungleboyj | xyang: So, right now we have doc/source/sample_config.rst that creates a web page that has the sample.conf file in it. | 16:28 |
| jungleboyj | xyang: That isn't what you were looking for though. Right? | 16:29 |
| xyang | jungleboyj: we just found out this is missing: https://github.com/openstack/openstack-manuals/blob/master/doc/config-reference/source/block-storage/drivers/emc-vmax-driver.rst | 16:30 |
| xyang | jungleboyj: so I don't know where it goes | 16:31 |
| openstackgerrit | John Griffith proposed openstack/cinder master: Add ability to extend 'in-use' volume https://review.openstack.org/454287 | 16:31 |
| xyang | jungleboyj: I assume it should be moved to cinder doc folder eventually. it has all driver docs | 16:31 |
| xyang | jungleboyj: https://github.com/openstack/openstack-manuals/tree/stable/ocata/doc/config-reference | 16:32 |
| jungleboyj | Yeah, smcginnis pointed that out to Doug and it sounds like I will need to move that over | 16:32 |
| xyang | jungleboyj: it is still in ocata | 16:32 |
| xyang | jungleboyj: ok | 16:32 |
| jungleboyj | Right, they deleted all that stuff from Pike. | 16:32 |
| xyang | jungleboyj: sure | 16:33 |
| jungleboyj | Thanks for finding that. Let me figure out how to integrate that into our local tree and push up a patch. | 16:33 |
| xyang | jungleboyj: thanks! | 16:33 |
| jgriffith | xyang You're KILLING me!! :) | 16:34 |
| jungleboyj | xyang: Welcome. | 16:34 |
| jgriffith | xyang we're going to have to put a LOC limit on your patches from now on :) | 16:34 |
| xyang | jgriffith: what did I do:) | 16:34 |
| jungleboyj | jgriffith: I feel your pain? ;-) | 16:34 |
| jungleboyj | jgriffith: We have been trying to do that for years! | 16:34 |
| xyang | jgriffith: I don't have big LOC patches any more:) | 16:34 |
| jgriffith | then why am I reviewing one right now :) | 16:34 |
| jgriffith | https://review.openstack.org/#/c/477688/13 | 16:35 |
| xyang | jgriffith: oh that one, please review it! | 16:35 |
| jgriffith | haha! | 16:35 |
| jgriffith | Nope, you said you didn't have any | 16:35 |
| xyang | jgriffith: that's a small one:) | 16:35 |
| jgriffith | I'm going to play PingPong :) | 16:35 |
| xyang | jgriffith: :) | 16:35 |
| *** ChubYann has joined #openstack-cinder | 16:36 | |
| jgriffith | xyang you get the "best Cinder contributor" award by the way, for being so thorough and updating all of the drivers!! | 16:37 |
| xyang | jgriffith: I have to because otherwise I can't pass Jenkins:) | 16:37 |
| jgriffith | lots of people would've cheated and made a kwargs or something there | 16:37 |
| xyang | jgriffith: oh, I didn't think of that. will try next time:) | 16:38 |
| jgriffith | xyang Ohhh, so what you're saying is you're NOT really that good, you just didn't have any choice :) | 16:38 |
| jgriffith | haha! | 16:38 |
| xyang | jgriffith: :) | 16:38 |
| jgriffith | xyang don't you go changing | 16:38 |
| jungleboyj | :-) | 16:38 |
| jgriffith | it's not as bad as I thought to review given that all the drivers are pretty consistent here | 16:39 |
| xyang | jgriffith: see, I told you it's a small patch:) | 16:40 |
| jgriffith | I know who to ping now when I need help with unit tests :) | 16:40 |
| xyang | :) | 16:40 |
| xyang | jgriffith: thanks! | 16:41 |
| *** itlinux has joined #openstack-cinder | 17:09 | |
| ericyoung | Anyone have some time to review a ScaleIO driver patch... https://review.openstack.org/#/c/478644/ | 17:13 |
| jungleboyj | ericyoung: I will try to take a look later. | 17:14 |
| ericyoung | jungleboyj: much appreciated | 17:14 |
| ericyoung | jungleboyj: quite a bit of cleanup in that patch...feel free to ask any questions | 17:15 |
| jungleboyj | ericyoung: Will do. | 17:17 |
| *** lpetrut has joined #openstack-cinder | 17:19 | |
| openstackgerrit | Rawan Herzallah proposed openstack/os-brick master: Adding NVMEoF for initiator CLI https://review.openstack.org/482642 | 17:26 |
| *** gheorghe_ has joined #openstack-cinder | 17:29 | |
| *** edmondsw has joined #openstack-cinder | 17:35 | |
| *** eharney has joined #openstack-cinder | 17:40 | |
| mriedem | smcginnis: ildikov: jgriffith: just reading scrollback, i still don't think we want to support extending the size of a reserved volume, | 17:42 |
| mriedem | we know that's going to be unpredictable on the nova side, so why even bother risking it? | 17:42 |
| smcginnis | mriedem: Yep, i think that's where we ended up. | 17:42 |
| mriedem | ok, whew | 17:43 |
| jgriffith | mriedem correct, that's what we decided | 17:43 |
| jgriffith | mriedem I just added some notes in the docs to make it clear and gave it a +2 | 17:43 |
| *** chhavi has quit IRC | 17:43 | |
| smcginnis | mriedem: jgriffith Just looking at that latest one. mriedem want to take a look before I +W? | 17:43 |
| mriedem | let me take a glance quick | 17:44 |
| hemna | someone wants to be able to extend a reserved volume ?! | 17:44 |
| smcginnis | hemna: No, not really. | 17:45 |
| hemna | ok phew | 17:45 |
| *** zzzeek_ has quit IRC | 17:46 | |
| *** zzzeek_ has joined #openstack-cinder | 17:49 | |
| mriedem | jgriffith: lgtm | 17:50 |
| mriedem | +1 | 17:50 |
| *** lpetrut has quit IRC | 17:51 | |
| *** raunak has joined #openstack-cinder | 18:01 | |
| *** raunak has quit IRC | 18:02 | |
| *** eharney has quit IRC | 18:05 | |
| *** diablo_rojo has quit IRC | 18:06 | |
| *** lpetrut has joined #openstack-cinder | 18:13 | |
| *** eharney has joined #openstack-cinder | 18:18 | |
| *** edmondsw_ has joined #openstack-cinder | 18:23 | |
| *** tesseract has quit IRC | 18:25 | |
| *** edmondsw has quit IRC | 18:25 | |
| *** lpetrut has quit IRC | 18:47 | |
| *** crushil has quit IRC | 18:49 | |
| *** crushil has joined #openstack-cinder | 18:50 | |
| *** nicolasbock has quit IRC | 18:50 | |
| patrickeast | would be curious to hear if others have opinions on https://review.openstack.org/#/c/479392 for where we would prefer to turn on cache limits for cinder testing and stuff | 18:57 |
| openstackgerrit | Patrick East proposed openstack/cinder master: Add support for shared "backend_defaults" config https://review.openstack.org/335299 | 18:58 |
| *** sdague has quit IRC | 19:00 | |
| smcginnis | patrickeast: So we should all switch to systemd because it's closer to real deployment, but we shouldn't use realistic limits because why do developers need that? :| | 19:02 |
| patrickeast | hah | 19:04 |
| patrickeast | smcginnis: i guess so, yea | 19:04 |
| *** arnewiebalck_ has joined #openstack-cinder | 19:06 | |
| geguileo | smcginnis: eharney I'll post a solution to the os-brick problem by tomorrow | 19:09 |
| smcginnis | geguileo: Great! Does that mean you've found the root problem? | 19:10 |
| geguileo | smcginnis: I have a poc working now, but I want to make it more robust, add unit tests, and test it with multipath as well | 19:10 |
| geguileo | smcginnis: yes | 19:10 |
| smcginnis | geguileo: Awesome, thanks for doing that. | 19:10 |
| *** mvk has quit IRC | 19:11 | |
| *** itlinux has quit IRC | 19:47 | |
| *** diablo_rojo has joined #openstack-cinder | 19:56 | |
| *** diablo_rojo has quit IRC | 20:05 | |
| *** diablo_rojo has joined #openstack-cinder | 20:05 | |
| *** edmondsw_ is now known as edmondsw | 20:07 | |
| *** itlinux has joined #openstack-cinder | 20:10 | |
| *** Nel1x has joined #openstack-cinder | 20:20 | |
| *** MVenesio has quit IRC | 20:25 | |
| *** adisky__ has quit IRC | 20:36 | |
| *** arnewiebalck_ has quit IRC | 20:44 | |
| *** zzzeek_ has quit IRC | 20:49 | |
| *** catintheroof has quit IRC | 21:04 | |
| *** jmlowe has joined #openstack-cinder | 21:05 | |
| openstackgerrit | Eric Harney proposed openstack/cinder master: Fix python 3.6 escape char warnings in strings https://review.openstack.org/482704 | 21:05 |
| smcginnis | eharney: 3.6?! :) | 21:05 |
| eharney | smcginnis: i upgraded to Fedora 26 today, now my unit tests are noisy :) | 21:06 |
| smcginnis | eharney: You have to put that on docstrings even? | 21:07 |
| smcginnis | eharney: Does sphynx like that? | 21:07 |
| eharney | smcginnis: yeah, it's a string issue, not a regex issue | 21:07 |
| eharney | i don't know why it wouldn't, the "r" just affects the parsing, they still end up as strings | 21:08 |
| smcginnis | eharney: Not sure how that parses source. | 21:08 |
| eharney | smcginnis: converting them to raw seemed easier than having to redraw ascii diagrams that used backslashes for arrows :) | 21:09 |
| smcginnis | bleh | 21:09 |
| *** zzzeek_ has joined #openstack-cinder | 21:11 | |
| *** dustins has quit IRC | 21:22 | |
| *** pcaruana has quit IRC | 21:23 | |
| *** abishop_ has quit IRC | 21:23 | |
| diablo_rojo | smcginnis, so, been looking at condensing configuration options for a large chunk of today, and man...that will take a very long time. Its very very tedious. | 21:37 |
| smcginnis | diablo_rojo: No doubt. | 21:37 |
| diablo_rojo | Also, it occurs to me that in order to make a set that everyone can use, we will have more options during the deprecation period.. | 21:38 |
| openstackgerrit | Eric Harney proposed openstack/cinder master: Dell SC: Fix python 3.6 'async' keyword warning https://review.openstack.org/482713 | 21:39 |
| patrickeast | what was the motivation for condensing the options again? | 21:39 |
| diablo_rojo | smcginnis, if you want me to put up a patch to start condensing them- even a few options at a time it's going to take some time. | 21:39 |
| diablo_rojo | patrickeast, uhhh, I don't remember who lead the charge.. | 21:39 |
| diablo_rojo | jgriffith, ? | 21:40 |
| diablo_rojo | geguileo, was in the conversation too if I remember | 21:40 |
| smcginnis | diablo_rojo: Was the diea behind that that we have so many options that are similar but different for each driver, so collapse some of them into some more commonly applicable ones? | 21:40 |
| smcginnis | patrickeast: I think it was out of the DISCO driver trying to add new san_ip, san_user, etc. options. | 21:41 |
| diablo_rojo | I think so? But wasn't that kind of the original intention with the SAN ones that don't get used almost ever? | 21:41 |
| patrickeast | smcginnis: gotcha | 21:41 |
| patrickeast | so whats the benefit of doing all that? | 21:41 |
| eharney | yes that was the idea | 21:41 |
| patrickeast | or more importantly do i need to change my drivers? :D | 21:41 |
| patrickeast | like... is this to save us headaches? simplify the deployments? | 21:42 |
| eharney | main benefit IMO is that it's not a giant pain for deployers to figure out how to write a correct config file | 21:42 |
| smcginnis | eharney: +1 | 21:43 |
| diablo_rojo | That makes sense. | 21:43 |
| patrickeast | is it really going to help though? like.. if we have one thing called "management_ip" or something (an an example) | 21:43 |
| smcginnis | This seems like it would be a really long term thing before everyone would be able to really take advantage of it, but if we could semi-standardize one some things, it would make it a little simpler. | 21:43 |
| patrickeast | isn't it the same amount of documentation required to read for a deployer to figure out what that does for their storage array versus something called "pure_management_vip" | 21:44 |
| patrickeast | or whatever | 21:44 |
| diablo_rojo | Should we agree on a general set of options and how they are defined and then collapse? Cause, looking at ALL THE OPTIOSN FOR ALL THE THINGS is making me want to claw my eyes out. | 21:44 |
| eharney | it's not the same amount of reading if they have two different arrays | 21:44 |
| smcginnis | patrickeast: Well, if everyone used management_ip, then if you've had some experience, you wouldn't need to look up specific options each time you work with a driver. | 21:44 |
| smcginnis | At least that's the hope. | 21:45 |
| patrickeast | mm i guess | 21:45 |
| patrickeast | i dunno, there is enough subtle differences that i'm skeptical it would really be much less confusing | 21:45 |
| eharney | this consolidation will also help with efforts where we (someone) have to wire in every new option to tripleo for deployment tools to work for more drivers | 21:45 |
| patrickeast | look at like "san_ip" today | 21:45 |
| patrickeast | its not even an ip on the san for us | 21:45 |
| jgriffith | patrickeast diablo_rojo I was bitching because somebody had a patch up for their driver that had "device_xyz_api_endpoint", "device_xyz_admin_login"... etc etc | 21:45 |
| jgriffith | I suggested they just use san_ip, san_login etc | 21:46 |
| jgriffith | the response was "it's not really a san" | 21:46 |
| patrickeast | haha | 21:46 |
| jgriffith | "I don't know what those are" | 21:46 |
| jgriffith | blah blah blah | 21:46 |
| smcginnis | Or "it doesn | 21:46 |
| smcginnis | t use the san protocol", whatever that means. | 21:46 |
| jgriffith | then I went on a rant on how f'ing ridiculous it is that every driver has it's own custom variable/conf entries and exception entries that are 99.99999% exactly the same | 21:47 |
| *** eharney has quit IRC | 21:47 | |
| diablo_rojo | jgriffith, lol noted :) | 21:47 |
| jgriffith | So then I said in the meeting: "generate a sample-conf and open the file, take a look and tell me how to sort through that" | 21:47 |
| jgriffith | diablo_rojo said something like "OMG" | 21:47 |
| jgriffith | or "that's terrible" or "I"m blind!!!!" | 21:48 |
| jgriffith | can't remember exactly | 21:48 |
| jgriffith | somebody said "we should fix that" I laughed, and said "yes we should" | 21:48 |
| jgriffith | that's about all I remember ;) | 21:48 |
| jgriffith | diablo_rojo ok... so now seriously :) | 21:49 |
| diablo_rojo | jgriffith, sounds about right. I remember being led into volunteering. Not sure exactly why I agreed to this torture. | 21:49 |
| diablo_rojo | Lol | 21:49 |
| jgriffith | diablo_rojo yes, the problem with doing something like this and I do believe we discussed it was that it sucks because if you follow deprecation rules etc that means you implement MORE duplicates to remove duplicates which is stupid | 21:49 |
| *** Apoorva has quit IRC | 21:50 | |
| jgriffith | I thought we came up with an idea of trying to just consolidate some of the most obvious ones or something, but I don't know | 21:50 |
| patrickeast | that makes sense, but i guess i'm not really sure that fewer abstract options are better than more specific-to-backend ones | 21:50 |
| jgriffith | patrickeast seems about right coming from YOU!!! | 21:50 |
| jgriffith | Just kidding | 21:50 |
| patrickeast | haha | 21:50 |
| patrickeast | hey, we use like 50/50 shared versus pure_* options :P | 21:50 |
| jgriffith | patrickeast haha | 21:51 |
| patrickeast | and the problem i always have is with the vague shared ones | 21:51 |
| jgriffith | well then you're in the top 90 percentile | 21:51 |
| patrickeast | that no one can understand | 21:51 |
| jgriffith | yeah, I get it; it's just kinda sad when you look at our conf-sample | 21:51 |
| patrickeast | the ones that are like "pure_do_the_thing = true" are way more self documenting | 21:51 |
| jgriffith | it's not really a sample for anything other than "look how messy this is" | 21:51 |
| smcginnis | diablo_rojo: Maybe hold off on doing anything. This seems like something we should have a PTG session on. | 21:51 |
| jgriffith | hey.. here's another idea | 21:51 |
| smcginnis | Or a Friday afternoon rant session. | 21:52 |
| patrickeast | sounds like maybe we just need to organize it better? | 21:52 |
| jgriffith | diablo_rojo I'm not saying you should do this (don't want you blaming me later).... | 21:52 |
| jgriffith | what if we tuned up a config-generator that you could specify what backend you wanted? | 21:52 |
| jgriffith | and what options were in your cloud? | 21:52 |
| jgriffith | so like if you don't run c-backup or swift you don't get those options, and if you run pure you don't get SolidFire options? | 21:53 |
| jgriffith | Or people can just use grep | 21:53 |
| jgriffith | and their eyes | 21:53 |
| jgriffith | or whatever | 21:53 |
| * jgriffith shuts up now | 21:53 | |
| smcginnis | jgriffith: Like this? https://review.openstack.org/#/c/474307/ | 21:53 |
| patrickeast | haha yea i was just looking for that patch | 21:54 |
| patrickeast | if we had them organized somehow it wouldn't be so scary | 21:54 |
| jgriffith | OH, well there ya go! | 21:54 |
| patrickeast | just group the backend options | 21:54 |
| jgriffith | diablo_rojo is so quick she actually implemented it 2 weeks before I mentioned it :) | 21:54 |
| jgriffith | diablo_rojo I said "DONT GO DO IT YET" | 21:54 |
| jgriffith | geesh!! | 21:54 |
| jgriffith | :) | 21:54 |
| jgriffith | Oh, that was Doug | 21:55 |
| jgriffith | never mind | 21:55 |
| diablo_rojo | jgriffith, I will take credit for it since I did +1 ;) | 21:55 |
| jgriffith | diablo_rojo works for me :) | 21:55 |
| diablo_rojo | Moral support is worth a lot you know :) | 21:55 |
| jgriffith | sometimes it's worth *everything* | 21:56 |
| openstackgerrit | John Griffith proposed openstack/cinder master: organize options based on backend drivers https://review.openstack.org/474307 | 21:56 |
| diablo_rojo | He needs some +1 and +2 love | 21:56 |
| diablo_rojo | jgriffith, was that on purpose? lol | 21:56 |
| jgriffith | No | 21:56 |
| jgriffith | :) | 21:56 |
| jgriffith | but it won't hurt anything | 21:57 |
| jgriffith | I hope | 21:57 |
| jgriffith | diablo_rojo he had a -1 on it and jenkins fail | 21:57 |
| diablo_rojo | I thought not. Now it's gonna look kinda weird when you +2 | 21:57 |
| *** jmlowe has quit IRC | 21:57 | |
| jgriffith | I was looking to see if the patch you mentioned that landed fixes jenkins up | 21:57 |
| diablo_rojo | smcginnis, sounds good. I can add it to the ptg agenda. | 21:57 |
| jgriffith | diablo_rojo why would that look weird? | 21:57 |
| jgriffith | I just hit rebase on it | 21:58 |
| jgriffith | if that's weird I'm in super big trouble :) | 21:58 |
| diablo_rojo | I dunno, just a passing thought. Nevermind me :) | 21:58 |
| diablo_rojo | I guess it won't | 21:58 |
| jgriffith | I'm also assuming doug ws going to come back to do something else with it | 21:58 |
| diablo_rojo | Lol, blame the config options. My brain is becoming mush. | 21:58 |
| diablo_rojo | Also booking flights for the other chunk of today.. | 21:59 |
| jgriffith | config options will most certainly ruin your brain | 21:59 |
| jgriffith | maybe that can be my excuse :) | 21:59 |
| diablo_rojo | jgriffith, I just need one more config option. I CAN STOP WHENEVER I WANT | 21:59 |
| jgriffith | ha!!!!! | 21:59 |
| jgriffith | Nice! | 21:59 |
| dfisher | ini-style config is much better than ifdef's all over .c/.h code. | 21:59 |
| jgriffith | T-shirts | 21:59 |
| jgriffith | dfisher says who? | 21:59 |
| dfisher | me. | 21:59 |
| dfisher | loudly. | 22:00 |
| jgriffith | bully for you | 22:00 |
| dfisher | that's right. | 22:00 |
| diablo_rojo | jgriffith, +1 to tshirts with that and the cinder logo | 22:00 |
| jgriffith | :) | 22:00 |
| diablo_rojo | *mascot | 22:00 |
| jgriffith | dfisher I'm just kidding by the way | 22:00 |
| dfisher | i figured ;) | 22:00 |
| dfisher | building Erlang for SPARC is the opposite of fun. | 22:00 |
| * jgriffith resists tempation to comment on building any C program on SPARC | 22:01 | |
| *** jmlowe has joined #openstack-cinder | 22:01 | |
| jgriffith | it was nice when it was running :) | 22:01 |
| dfisher | pkg install gcc. gcc foo.c ;) | 22:01 |
| dfisher | done. | 22:01 |
| jgriffith | only thing worse was AIX and old Smitty | 22:01 |
| jgriffith | dfisher well.. yeah, depending on what sort of libs foo.c pulled in, and if you used so's | 22:02 |
| dfisher | ;) | 22:02 |
| dfisher | it's better now. | 22:02 |
| dfisher | but, I understand. | 22:02 |
| jgriffith | BUT to be fair, this was 15 years ago and I worked on Linux and HPUX | 22:02 |
| jgriffith | so anything else was just "hard" | 22:02 |
| jgriffith | I'm sure everybody running SPARC said the same about crappy HPUX | 22:03 |
| dfisher | we still say it. | 22:03 |
| ildikov | jgriffith: what's the problem with AIX? :) | 22:03 |
| jgriffith | etc etc | 22:03 |
| jgriffith | haha!! | 22:03 |
| jgriffith | ildikov I knew I was going to catch it from somebody! | 22:03 |
| jgriffith | :) | 22:03 |
| dfisher | i know this is the wrong community for this, but it's really too bad Solaris isn't more prevalent. It's really really good. :/ | 22:03 |
| jgriffith | Nothing, TBC nothing at all! | 22:03 |
| jgriffith | except Smitty was kind of an awkard sysadmin tool IMO | 22:04 |
| ildikov | jgriffith: my pleasure then :) | 22:04 |
| dfisher | alas, Oracle had to go be dbags and re-close it. | 22:04 |
| openstackgerrit | Helen Walsh proposed openstack/cinder master: VMAX driver - seamless upgrade from SMI-S to REST https://review.openstack.org/482138 | 22:04 |
| jgriffith | but again, probably because I was coming from my sheltered HPUX world at the time | 22:04 |
| *** Apoorva has joined #openstack-cinder | 22:04 | |
| *** felipemonteiro__ has quit IRC | 22:04 | |
| jgriffith | I just remember being at a customer site in their DC at 3 in the morning watching that stupid running man graphic over and over and over, only to have it fail | 22:05 |
| jgriffith | kept thinking they should have made him explode or something when a task failed | 22:05 |
| jgriffith | or run into a wall | 22:05 |
| ildikov | I think running over and over again and not to explode when the task failed was kind of a bigger torture :) | 22:07 |
| dfisher | or trip over Clippy? | 22:08 |
| *** jmlowe has quit IRC | 22:09 | |
| ildikov | oooh the good old days, when Clippy started to snore on the projector one of my university classes... :) | 22:10 |
| ildikov | that teacher was hopeless... | 22:10 |
| *** dave-mccowan has quit IRC | 22:14 | |
| *** markvoelker has quit IRC | 22:16 | |
| *** markvoelker has joined #openstack-cinder | 22:17 | |
| jungleboyj | diablo_rojo: jgriffith Just finally catching up with the backlog here. | 22:17 |
| jungleboyj | diablo_rojo: jgriffith The direction we were looking to go was to take existing config options and try to standardize around them so that we could not have to add a bunch more during the deprecation period. | 22:18 |
| *** gnufied has quit IRC | 22:18 | |
| jungleboyj | Do it little by little starting with easy/obvious ones like san_ip. | 22:18 |
| jungleboyj | Anyway, if we can go into the PTG with some idea of what those common options are so that it is not just a bitch session it would be good. I really think we need to try to slow our config option sprawl though. | 22:19 |
| ildikov | jungleboyj: that and the proposal about adding only those to the config file which driver is configured could result in a kind of condensed output by the end IMHO | 22:21 |
| jungleboyj | ildikov: Agreed. Being able to create a more focused sample config file would also be a good improvement. | 22:23 |
| ildikov | jungleboyj: +1 | 22:24 |
| jungleboyj | ildikov: Also hoping to help the situation somewhat with the documentation improvements. | 22:25 |
| ildikov | jungleboyj: yeah, that's a tough one | 22:26 |
| *** edmondsw has quit IRC | 22:26 | |
| jungleboyj | ildikov: How so> | 22:30 |
| *** dr_gogeta86 has quit IRC | 22:30 | |
| ildikov | jungleboyj: that's what comes to my mind when I think about config options and docs at the same time :) | 22:30 |
| ildikov | jungleboyj: spent a little time with the scripts back at the time that generate the Config Guide | 22:31 |
| ildikov | jungleboyj: I still have all my respect to the team to put all that automation together | 22:31 |
| *** gnufied has joined #openstack-cinder | 22:31 | |
| jungleboyj | ildikov: Well, I am just hoping to be able to keep things equally good as I bring the docs over and then hopefully improve it over time. | 22:34 |
| ildikov | jungleboyj: it will be all good | 22:34 |
| jungleboyj | ildikov: Thanks. Keep sending me moral support. :-) | 22:35 |
| ildikov | jungleboyj: moving the docs over has many advantages, so keep those in mind :) | 22:36 |
| jungleboyj | ildikov: Thanks. | 22:36 |
| ildikov | jungleboyj: you can sort out the rest step by step | 22:36 |
| jungleboyj | ildikov: That is the plan. :-) | 22:36 |
| ildikov | jungleboyj: sounds good then :) | 22:37 |
| *** dfisher has quit IRC | 22:44 | |
| *** itlinux has quit IRC | 23:10 | |
| openstackgerrit | Merged openstack/cinder master: Windows iSCSI: use configured iSCSI portals https://review.openstack.org/464659 | 23:28 |
| openstackgerrit | Merged openstack/cinder master: Allow extension of volumes with snapshots for SMBfs driver https://review.openstack.org/462902 | 23:30 |
| *** diablo_rojo has quit IRC | 23:46 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!