*** Masahiro has joined #openstack-swift | 00:01 | |
*** Masahiro has quit IRC | 00:05 | |
*** rdaly2 has quit IRC | 00:17 | |
notmyname | gyee: looking | 00:25 |
---|---|---|
gyee_ | notmyname, thanks | 00:25 |
notmyname | gyee: I'm pretty sure there's a way, but I haven't looked at that section of code recently enough to remember. checking now | 00:25 |
gyee_ | my understanding is that it requires db hack | 00:25 |
gyee_ | like manually reverse the status | 00:26 |
notmyname | so far I'm reaching that same conclusion | 00:28 |
gyee_ | notmyname, I was looking at this doc http://docs.openstack.org/developer/swift/overview_reaper.html | 00:28 |
gyee_ | not sure if its up-to-date though | 00:28 |
*** jwang__ has quit IRC | 00:28 | |
notmyname | (for others, we're talking about how to undelete an account before the reaper picks it up) | 00:28 |
*** nellysmitt has joined #openstack-swift | 00:29 | |
gyee_ | notmyname, cool, thanks for the help! | 00:30 |
notmyname | gyee_: yup. that's the current truth of things. that doc is correct | 00:34 |
*** nellysmitt has quit IRC | 00:34 | |
gyee_ | notmyname, cool, thanks! | 00:35 |
notmyname | gyee_: that being said, I'd love to see a patch for that utility | 00:36 |
gyee_ | me too :) | 00:37 |
gyee_ | I'll see what I can do | 00:37 |
notmyname | gyee_: yay! | 00:37 |
*** dmorita has joined #openstack-swift | 00:50 | |
*** jwang__ has joined #openstack-swift | 00:55 | |
*** Masahiro has joined #openstack-swift | 00:55 | |
*** addnull has joined #openstack-swift | 01:03 | |
*** Masahiro has quit IRC | 01:07 | |
*** tellesnobrega_ has quit IRC | 01:10 | |
*** Masahiro has joined #openstack-swift | 01:10 | |
openstackgerrit | Clay Gerrard proposed openstack/swift: Fix reclaim on deleted containers https://review.openstack.org/138924 | 01:11 |
*** dmsimard is now known as dmsimard_away | 01:12 | |
*** rmcall has quit IRC | 01:12 | |
*** occupant has joined #openstack-swift | 01:27 | |
*** gyee_ has quit IRC | 01:28 | |
*** tellesnobrega_ has joined #openstack-swift | 01:32 | |
*** bill_az has quit IRC | 01:33 | |
*** nosnos has joined #openstack-swift | 01:52 | |
*** sluo_wfh has joined #openstack-swift | 01:58 | |
*** sluo_wfh has quit IRC | 02:03 | |
*** haomaiwa_ has joined #openstack-swift | 02:11 | |
*** Masahiro has quit IRC | 02:20 | |
*** Masahiro has joined #openstack-swift | 02:23 | |
*** nellysmitt has joined #openstack-swift | 02:30 | |
*** sluo_wfh has joined #openstack-swift | 02:30 | |
*** nellysmitt has quit IRC | 02:35 | |
*** Masahiro has quit IRC | 02:38 | |
*** Masahiro has joined #openstack-swift | 02:41 | |
openstackgerrit | Merged openstack/swift: Clean up empty account and container partitions directories. https://review.openstack.org/138524 | 02:44 |
*** cppforlife has joined #openstack-swift | 02:52 | |
mattoliverau | Sigh, you know your unit test-fu is a little off when you write a unit test to actally tests concurrent reads, which actually seems to succeeds but almost takes down my dev desktop when the test method runs! :P Might test it somewhere else before I take down infra and the swift stack bot :P | 03:02 |
mattoliverau | in other words.. back to debugging :P | 03:02 |
*** tellesnobrega_ has quit IRC | 03:10 | |
*** Masahiro has quit IRC | 03:18 | |
*** Masahiro has joined #openstack-swift | 03:19 | |
*** rdaly2 has joined #openstack-swift | 03:20 | |
*** jwang__ has quit IRC | 03:22 | |
*** nosnos has quit IRC | 03:23 | |
*** rdaly2 has quit IRC | 03:25 | |
*** Masahiro has quit IRC | 03:27 | |
*** Masahiro has joined #openstack-swift | 03:30 | |
*** rmcall has joined #openstack-swift | 03:33 | |
*** rmcall has quit IRC | 03:38 | |
*** bkopilov has quit IRC | 03:39 | |
*** nosnos has joined #openstack-swift | 04:16 | |
*** sandywalsh_ has quit IRC | 04:18 | |
*** SkyRocknRoll has joined #openstack-swift | 04:23 | |
*** echevemaster has quit IRC | 04:30 | |
*** nellysmitt has joined #openstack-swift | 04:31 | |
*** Masahiro has quit IRC | 04:31 | |
*** nellysmitt has quit IRC | 04:36 | |
*** Masahiro has joined #openstack-swift | 04:42 | |
*** rmcall has joined #openstack-swift | 04:44 | |
*** rmcall_ has joined #openstack-swift | 04:47 | |
*** rmcall has quit IRC | 04:48 | |
*** rmcall has joined #openstack-swift | 04:51 | |
*** rmcall_ has quit IRC | 04:51 | |
*** kopparam has joined #openstack-swift | 04:53 | |
*** bkopilov has joined #openstack-swift | 04:58 | |
*** kopparam has quit IRC | 05:03 | |
*** Masahiro has quit IRC | 05:06 | |
*** erlon has quit IRC | 05:09 | |
*** erlon has joined #openstack-swift | 05:09 | |
*** erlon has quit IRC | 05:09 | |
*** erlon has joined #openstack-swift | 05:09 | |
*** kopparam has joined #openstack-swift | 05:16 | |
*** Masahiro has joined #openstack-swift | 05:16 | |
openstackgerrit | Matthew Oliver proposed openstack/swift: Add concurrent reads option to proxy https://review.openstack.org/117710 | 05:21 |
*** dmorita has quit IRC | 05:21 | |
*** rdaly2 has joined #openstack-swift | 05:22 | |
*** addnull has quit IRC | 05:23 | |
*** ahonda has quit IRC | 05:23 | |
*** ahonda has joined #openstack-swift | 05:24 | |
*** zaitcev has quit IRC | 05:26 | |
*** rdaly2 has quit IRC | 05:26 | |
*** Masahiro has quit IRC | 05:26 | |
*** Masahiro has joined #openstack-swift | 05:32 | |
*** sungju has quit IRC | 05:50 | |
*** Masahiro has quit IRC | 05:52 | |
*** Masahiro has joined #openstack-swift | 05:55 | |
*** addnull has joined #openstack-swift | 05:59 | |
*** kopparam has quit IRC | 06:02 | |
*** kopparam has joined #openstack-swift | 06:02 | |
*** haomaiwa_ has quit IRC | 06:03 | |
*** Masahiro has quit IRC | 06:05 | |
*** Masahiro has joined #openstack-swift | 06:05 | |
mattoliverau | Well, I'm calling it a night. Sleep well swiftland.. or as the case may be have a great day to those of you who are just waking up! | 06:06 |
*** nshaikh has joined #openstack-swift | 06:11 | |
*** nellysmitt has joined #openstack-swift | 06:32 | |
*** nellysmitt has quit IRC | 06:37 | |
*** kopparam has quit IRC | 06:39 | |
*** kopparam has joined #openstack-swift | 06:39 | |
*** addnull has quit IRC | 06:41 | |
*** k4n0 has joined #openstack-swift | 06:54 | |
*** addnull has joined #openstack-swift | 07:01 | |
*** haomaiwa_ has joined #openstack-swift | 07:05 | |
*** rmcall has quit IRC | 07:07 | |
*** addnull has quit IRC | 07:16 | |
*** addnull has joined #openstack-swift | 07:16 | |
*** rdaly2 has joined #openstack-swift | 07:23 | |
*** rdaly2 has quit IRC | 07:28 | |
*** Masahiro has quit IRC | 07:38 | |
*** Masahiro has joined #openstack-swift | 07:49 | |
*** addnull has quit IRC | 07:53 | |
*** addnull has joined #openstack-swift | 07:54 | |
*** rledisez has joined #openstack-swift | 08:08 | |
*** nellysmitt has joined #openstack-swift | 08:26 | |
*** geaaru has joined #openstack-swift | 08:36 | |
*** jordanP has joined #openstack-swift | 08:53 | |
*** kopparam has quit IRC | 09:01 | |
*** jistr has joined #openstack-swift | 09:05 | |
*** jordanP has quit IRC | 09:05 | |
openstackgerrit | Merged openstack/swift: Fix reclaim on deleted containers https://review.openstack.org/138924 | 09:10 |
*** jordanP has joined #openstack-swift | 09:17 | |
*** rdaly2 has joined #openstack-swift | 09:25 | |
*** nshaikh has quit IRC | 09:27 | |
*** rdaly2 has quit IRC | 09:29 | |
*** foexle has joined #openstack-swift | 09:31 | |
*** jistr is now known as jistr|trng | 09:35 | |
*** kopparam has joined #openstack-swift | 09:37 | |
*** nshaikh has joined #openstack-swift | 09:52 | |
*** addnull has quit IRC | 09:56 | |
*** rledisez has quit IRC | 09:59 | |
*** rledisez has joined #openstack-swift | 10:07 | |
*** Masahiro has quit IRC | 10:14 | |
*** tellesnobrega_ has joined #openstack-swift | 10:34 | |
openstackgerrit | Donagh McCabe proposed openstack/swift-specs: Minor updates to composite token spec https://review.openstack.org/138771 | 10:40 |
*** addnull has joined #openstack-swift | 10:44 | |
*** haomaiwa_ has quit IRC | 11:01 | |
*** tellesnobrega_ has quit IRC | 11:04 | |
*** Masahiro has joined #openstack-swift | 11:15 | |
*** Masahiro has quit IRC | 11:19 | |
*** kopparam has quit IRC | 11:24 | |
*** rdaly2 has joined #openstack-swift | 11:26 | |
*** tellesnobrega_ has joined #openstack-swift | 11:28 | |
*** kopparam has joined #openstack-swift | 11:30 | |
*** rdaly2 has quit IRC | 11:31 | |
*** acoles_away is now known as acoles | 11:35 | |
acoles | mattoliverau: hey matt, i used to almost overlap with you but now the clocks have changed it seems there is no chance | 11:37 |
acoles | mattoliverau: glad you din't wake early for the meeting! | 11:38 |
mattoliverau | acoles: yeah, perils of the opposite side of the globe, in spring (my time) I go forward and you guys go back, Autumn is the opposite. Making the gap either 2 hours closer or apart. | 11:47 |
mattoliverau | acoles: thanks for the review :) anyway sleep time :) | 11:48 |
acoles | mattoliverau: g'nite | 11:48 |
mattoliverau | Night (or I guess I should say g'day) | 11:49 |
*** aix has quit IRC | 11:59 | |
*** yuanz has quit IRC | 11:59 | |
*** yuanz has joined #openstack-swift | 12:00 | |
*** peluse_ has joined #openstack-swift | 12:00 | |
*** peluse has quit IRC | 12:00 | |
openstackgerrit | Hisashi Osanai proposed openstack/swift: Allow hostnames for nodes in Rings https://review.openstack.org/133155 | 12:33 |
*** kopparam_ has joined #openstack-swift | 12:42 | |
*** kopparam_ has quit IRC | 12:42 | |
*** kopparam has quit IRC | 12:44 | |
*** addnull has quit IRC | 13:08 | |
*** SkyRocknRoll has quit IRC | 13:11 | |
*** addnull has joined #openstack-swift | 13:12 | |
*** Masahiro has joined #openstack-swift | 13:17 | |
*** aix has joined #openstack-swift | 13:18 | |
mordred | an anybody point me to docs on how I get the info on a file I've uploaded to swift such that I can tell glance about it? | 13:20 |
*** aix has quit IRC | 13:20 | |
*** bkopilov has quit IRC | 13:20 | |
mordred | (specifically on rackspace, but I'd probably like to just learn more about the swift concepts at play anyway) | 13:20 |
*** Masahiro has quit IRC | 13:22 | |
*** nosnos has quit IRC | 13:27 | |
*** rdaly2 has joined #openstack-swift | 13:28 | |
*** bill_az has joined #openstack-swift | 13:29 | |
*** rdaly2 has quit IRC | 13:32 | |
*** bkopilov has joined #openstack-swift | 13:35 | |
*** tellesnobrega_ has quit IRC | 13:35 | |
*** nshaikh has quit IRC | 13:40 | |
*** addnull has quit IRC | 13:41 | |
*** tellesnobrega_ has joined #openstack-swift | 13:48 | |
*** bkopilov has quit IRC | 13:49 | |
*** ppai has joined #openstack-swift | 13:49 | |
*** addnull has joined #openstack-swift | 13:50 | |
*** addnull has quit IRC | 13:53 | |
*** miqui_ has joined #openstack-swift | 14:06 | |
*** jordanP has quit IRC | 14:07 | |
annegentle | mordred: so there's this https://developer.rackspace.com/docs/cloud-files/getting-started/#change-object-metadata but I am still looking for how to change object metadata | 14:12 |
*** jistr|trng has quit IRC | 14:13 | |
annegentle | mordred: then list it here http://docs.rackspace.com/files/api/v1/cf-devguide/content/HEAD_retrieveobjectmeta_v1__account___container___object__objectServicesOperations_d1e000.html | 14:14 |
annegentle | oh I meant "looking for how to view it" | 14:14 |
ahale | mordred: this kinda stuff, http://www.rackspace.com/knowledge_center/article/creating-an-ubuntu-1310-image-for-the-rackspace-open-cloud-0 ? | 14:15 |
ahale | i dont think theres much swift concepts in play in this :) | 14:16 |
mordred | ahale: yes - although I'd love if there was a version of that which used python-glanceclient - but that's super helpful | 14:17 |
mordred | ahale: I think the swift concept I'm missing is "so I've uploaded a file to swift, what url can I fetch it from?" | 14:17 |
*** aix has joined #openstack-swift | 14:18 | |
annegentle | mordred: you want temp url, https://developer.rackspace.com/docs/cloud-files/getting-started/#get-object | 14:19 |
ahale | aha yeah that'll be the container/obj name in the glance api reqs, i guess it already knows your tenant etc | 14:19 |
mordred | annegentle: I tried making one of those, I thought - but maybe with glance I don't actually need one of those, based on the curl examples | 14:20 |
*** ppai has quit IRC | 14:20 | |
*** jordanP has joined #openstack-swift | 14:20 | |
*** tellesnobrega_ has quit IRC | 14:21 | |
*** jistr has joined #openstack-swift | 14:22 | |
*** jistr is now known as jistr|trng | 14:22 | |
*** tellesnobrega_ has joined #openstack-swift | 14:24 | |
mordred | ahale: thank you VERY MUCH - that link was super helpful | 14:26 |
ahale | cool stuff :) | 14:26 |
mordred | ahale: for what it's worth, the python-glanceclient translation seems to be: | 14:26 |
mordred | glance --os-image-api-version=2 task-create --type=import --input='{"import_from": "images/test-monty-ubuntu.vhd", "image_properties" : {"name": "Test Monty Ubuntu"}}' | 14:26 |
ahale | oh neat, thats pretty straightforward | 14:27 |
mordred | it's still in "processing" state, so I'm going to assume that the json payload made sense :) | 14:27 |
*** k4n0 has quit IRC | 14:33 | |
*** miqui_ has quit IRC | 14:36 | |
*** tdasilva has joined #openstack-swift | 14:37 | |
*** tellesnobrega_ has quit IRC | 14:38 | |
*** miqui_ has joined #openstack-swift | 14:47 | |
*** sandywalsh has joined #openstack-swift | 14:56 | |
*** kopparam has joined #openstack-swift | 14:58 | |
*** lpabon has joined #openstack-swift | 14:58 | |
*** geaaru has quit IRC | 15:02 | |
*** kopparam has quit IRC | 15:04 | |
*** Masahiro has joined #openstack-swift | 15:06 | |
*** Masahiro has quit IRC | 15:11 | |
*** dmsimard_away is now known as dmsimard | 15:16 | |
*** pwehrle has joined #openstack-swift | 15:20 | |
*** pwehrle has quit IRC | 15:22 | |
*** kajinamit has joined #openstack-swift | 15:26 | |
kajinamit | Is there anybody available now? | 15:31 |
*** rdaly2 has joined #openstack-swift | 15:32 | |
kajinamit | I want somebody to review my patch(https://review.openstack.org/#/c/99598/) about bug in object-updater(https://bugs.launchpad.net/swift/+bug/1328735), which may cause dark object data in swift cluster not listed in container. | 15:33 |
openstackgerrit | Christian Schwede proposed openstack/python-swiftclient: Fix download redirect to stdout https://review.openstack.org/139091 | 15:35 |
openstackgerrit | Thiago da Silva proposed openstack/swift: added testing notes to the contributing doc https://review.openstack.org/138568 | 15:41 |
*** bkopilov has joined #openstack-swift | 15:41 | |
*** kajinamit has quit IRC | 15:43 | |
*** mahatic has joined #openstack-swift | 15:44 | |
*** aix has quit IRC | 15:49 | |
openstackgerrit | Alistair Coles proposed openstack/python-swiftclient: Fix cross account upload using --os-storage-url https://review.openstack.org/125759 | 15:57 |
*** aix has joined #openstack-swift | 16:01 | |
*** tellesnobrega_ has joined #openstack-swift | 16:07 | |
*** SkyRocknRoll has joined #openstack-swift | 16:08 | |
*** SkyRocknRoll has joined #openstack-swift | 16:08 | |
*** rdaly2_ has joined #openstack-swift | 16:12 | |
*** miqui__ has joined #openstack-swift | 16:15 | |
*** clds_ has joined #openstack-swift | 16:16 | |
*** zacksh has joined #openstack-swift | 16:17 | |
*** rdaly2 has quit IRC | 16:18 | |
*** jroll has quit IRC | 16:18 | |
*** miqui_ has quit IRC | 16:18 | |
*** jordanP has quit IRC | 16:18 | |
*** zacksh_ has quit IRC | 16:18 | |
*** ondergetekende has quit IRC | 16:18 | |
*** JelleB has quit IRC | 16:18 | |
*** mhu has quit IRC | 16:18 | |
*** aix has quit IRC | 16:18 | |
*** clds has quit IRC | 16:18 | |
*** jroll has joined #openstack-swift | 16:18 | |
*** jordanP has joined #openstack-swift | 16:19 | |
*** mhu has joined #openstack-swift | 16:19 | |
*** aix has joined #openstack-swift | 16:19 | |
*** ondergetekende has joined #openstack-swift | 16:19 | |
*** jroll has quit IRC | 16:19 | |
*** jroll has joined #openstack-swift | 16:19 | |
*** AbyssOne_ has joined #openstack-swift | 16:20 | |
*** flwang has quit IRC | 16:26 | |
*** nshaikh has joined #openstack-swift | 16:30 | |
*** silor has joined #openstack-swift | 16:33 | |
*** flwang has joined #openstack-swift | 16:39 | |
*** exploreshaifali has joined #openstack-swift | 16:49 | |
*** kopparam_ has joined #openstack-swift | 16:54 | |
*** Masahiro has joined #openstack-swift | 16:54 | |
*** Masahiro has quit IRC | 16:59 | |
*** nshaikh has quit IRC | 17:04 | |
*** tdasilva_ has joined #openstack-swift | 17:05 | |
*** gyee_ has joined #openstack-swift | 17:07 | |
*** tellesnobrega_ has quit IRC | 17:11 | |
*** tdasilva has quit IRC | 17:12 | |
*** foexle has quit IRC | 17:22 | |
*** rledisez has quit IRC | 17:24 | |
*** tellesnobrega_ has joined #openstack-swift | 17:25 | |
*** rdaly2_ has quit IRC | 17:30 | |
notmyname | good morning | 17:32 |
notmyname | clayg: around yet/ | 17:32 |
notmyname | ? | 17:32 |
*** annegent_ has joined #openstack-swift | 17:32 | |
*** aix has quit IRC | 17:35 | |
*** kopparam_ has quit IRC | 17:41 | |
*** jordanP has quit IRC | 17:44 | |
*** tellesnobrega_ has quit IRC | 17:44 | |
*** geaaru has joined #openstack-swift | 17:45 | |
*** annegent_ has quit IRC | 17:48 | |
*** rdaly2 has joined #openstack-swift | 17:48 | |
*** rdaly2 has quit IRC | 17:49 | |
*** rdaly2 has joined #openstack-swift | 17:49 | |
*** kopparam has joined #openstack-swift | 17:49 | |
*** jistr|trng has quit IRC | 17:51 | |
*** jwang__ has joined #openstack-swift | 17:55 | |
*** annegent_ has joined #openstack-swift | 17:57 | |
*** AbyssOne_ is now known as JelleB | 18:01 | |
*** rdaly2 has quit IRC | 18:04 | |
*** nshaikh has joined #openstack-swift | 18:18 | |
*** kopparam has quit IRC | 18:19 | |
ipolyzos | hello | 18:22 |
ipolyzos | try to use the keystone ec2-credentials-create but even though I provide the correct credential and produces an empty answer, could you please help me on how to proceed with identify potential problem that cause this ? | 18:24 |
*** jwang__ has quit IRC | 18:32 | |
notmyname | ipolyzos: I'm not familiar with that tool, but the "ec2" part of it makes me think you aren't using swift. if you're looking for some keystone-specific help, #openstack-keystone might be a good place. if you're looking for help with compute, #openstack-nova is where I'd go | 18:34 |
*** Masahiro has joined #openstack-swift | 18:43 | |
*** jwang__ has joined #openstack-swift | 18:43 | |
openstackgerrit | Merged openstack/swift: Fix container quota check during cross-account COPY https://review.openstack.org/136904 | 18:44 |
openstackgerrit | Merged openstack/swift: Raise ValueError for offset on Timestamp over limit https://review.openstack.org/138161 | 18:44 |
*** gyee_ has quit IRC | 18:45 | |
*** Masahiro has quit IRC | 18:48 | |
*** openstackgerrit has quit IRC | 18:50 | |
*** openstackgerrit has joined #openstack-swift | 18:50 | |
*** nellysmitt has quit IRC | 19:00 | |
*** tellesnobrega_ has joined #openstack-swift | 19:06 | |
*** pberis_ has joined #openstack-swift | 19:09 | |
*** annegent_ has quit IRC | 19:28 | |
ipolyzos | notmyname: thanks alot, I'm not very familiar also. Got it out was a config mistake in the env vars. Thanks alot in any case... | 19:30 |
openstackgerrit | Merged openstack/swift: added testing notes to the contributing doc https://review.openstack.org/138568 | 19:36 |
*** jwang__ has quit IRC | 19:37 | |
*** jwang__ has joined #openstack-swift | 19:38 | |
*** tellesnobrega_ has quit IRC | 19:38 | |
*** flwang has quit IRC | 19:40 | |
*** tellesnobrega_ has joined #openstack-swift | 19:42 | |
*** jbonjean has joined #openstack-swift | 19:45 | |
jbonjean | Hi, does anyone know why in the proxy server, connection timeout and refused are not considered as errors ? (so the error count not incremented ...) | 19:48 |
jbonjean | this mean the nodes that do not respond are never ignored | 19:49 |
*** flwang has joined #openstack-swift | 19:52 | |
*** rdaly2 has joined #openstack-swift | 19:52 | |
*** openstackgerrit has quit IRC | 20:04 | |
*** openstackgerrit has joined #openstack-swift | 20:04 | |
*** annegent_ has joined #openstack-swift | 20:05 | |
*** annegent_ has quit IRC | 20:08 | |
*** annegent_ has joined #openstack-swift | 20:12 | |
*** annegent_ has quit IRC | 20:21 | |
clayg | tdasilva_: thanks for updates on https://review.openstack.org/#/c/138568/ - totally got that bad boy merged! | 20:23 |
clayg | cschwede: looks like you need to six up https://review.openstack.org/#/c/139091/ - you around to do that or you need a push over? | 20:25 |
*** annegent_ has joined #openstack-swift | 20:26 | |
*** annegent_ has quit IRC | 20:26 | |
openstackgerrit | Alistair Coles proposed openstack/python-swiftclient: Fix cross account upload using --os-storage-url https://review.openstack.org/125759 | 20:30 |
*** acoles is now known as acoles_away | 20:32 | |
*** Masahiro has joined #openstack-swift | 20:32 | |
*** jwang__ has quit IRC | 20:35 | |
clayg | notmyname: this change got merged pretty fast (good) but I think was a sorta terrible bug - and I had the fix before even opening anything on launchpad - can you put on your community operator hat and figure out what we need to do re notification or worst case even backporting :\ | 20:36 |
clayg | redbo: dfg: ^ | 20:36 |
*** Masahiro has quit IRC | 20:37 | |
clayg | sorry, forgot the link! | 20:41 |
clayg | https://review.openstack.org/#/c/138924/ | 20:41 |
notmyname | clayg: I propose we use zaitcev's review comment as the changelog message: "Closed bug blah 'lift the fist of an angry god'" | 20:42 |
clayg | like? | 20:43 |
*** exploreshaifali has quit IRC | 20:43 | |
mattoliverau | Morning | 21:08 |
notmyname | mattoliverau: hi | 21:10 |
clayg | hi mattoliverau! | 21:11 |
*** tellesnobrega_ has quit IRC | 21:14 | |
*** silor has quit IRC | 21:14 | |
*** nshaikh has left #openstack-swift | 21:15 | |
tdasilva_ | clayg: hi, got time to talk about obj. versioning? | 21:17 |
clayg | tdasilva_: yeah ok | 21:20 |
tdasilva_ | clayg: just a quick question: in the commit message, you added a question about making version_writes a required middleware. I was thinking of removing the allow_versions option from the container server config. and just making the middleware optional | 21:22 |
clayg | tdasilva_: i have no idea what to do for upgrade path there :\ | 21:22 |
tdasilva_ | clayg: oh, yeah...I always forget to think of upgrades, good point..well, couldn't we just add to the release notes that the middleware should now be added to the pipeline for those deployments that had "allow_versions=true" | 21:24 |
*** SkyRocknRoll has quit IRC | 21:25 | |
clayg | tdasilva_: idk, in most cases that means the operator has to update a bunch of deployment tooling before they can update to the new version - and most of the time you want your deployment tooling to push out a single config that would be valid the old and new version | 21:26 |
clayg | so you have this change that requires you make different looking configs, you halt the upgrade train and update your deployment tools to push out a config that would also work with the new code, then you push out the new code, then you remove the cruft* | 21:26 |
clayg | * note the last part never happens | 21:26 |
clayg | if it's not a required middleware, then I can't put it in the pipeline? | 21:27 |
clayg | maybe if the proxy had an "make the versions works" flag that old code would simply ignore - then I could just turn it on in both places and after the upgrade everything still works? | 21:27 |
clayg | maybe the "make the versions works" flag is a noop if the middleware is alreayd in the pipeline or something and it defaults to true... i don't really know? | 21:28 |
tdasilva_ | clayg: how about checking in the proxy server where we add the required middleware to see if it's in the pipeline, if it's not and the old flag is still saying true, then auto add to the pipeline | 21:29 |
clayg | what old flag - the one in the container server? proxy may not have it | 21:29 |
tdasilva_ | oh, true | 21:30 |
clayg | notmyname: you screwed up when you put the version option in the container servers configs | 21:30 |
tdasilva_ | lol | 21:30 |
clayg | notmyname: next time use your time machine | 21:30 |
notmyname | ;-) | 21:30 |
notmyname | I screwed up when i didn't make it middleware ;-) | 21:30 |
tdasilva_ | notmyname: that's ok, gave opportunity to learn a bunch stuff :-) | 21:31 |
clayg | there's no problem in object storage that can't be solved with another piece of middleware exept too many middlewares? | 21:31 |
clayg | pretty sure someone said that | 21:31 |
notmyname | clayg: maybe one middleware that takes plugins? | 21:31 |
notmyname | maybe we could have a whole "openstack middleware" project! | 21:31 |
tdasilva_ | clayg: i will think a bit more about this problem...next question: you also added this note: "write new tests to verify correct handling metadata on backend COPYs" | 21:32 |
tdasilva_ | clayg: not sure what that means honestly | 21:32 |
tdasilva_ | clayg: what backend COPYs? | 21:34 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/swift: Updated from global requirements https://review.openstack.org/88736 | 21:34 |
clayg | tdasilva_: well the unittest's that did the real socket stuff caught the content-type thing, but I don't think they would have caught the same problem for x-object-meta | 21:35 |
clayg | i'm pretty sure your latest code get's metadata right just like content-type cause it uses subrequest or w/e - but it'd be nice to have a test that proves it | 21:35 |
clayg | I think "backedn copy" in this context was the GET/PUT request that deals with the current version before handling the PUT that triggered it (or something) | 21:36 |
clayg | tdasilva_: wasn't it like the PUT of the current version into the version backlog was getting headers that came along with the PUT that was trying to over-write it or something? | 21:36 |
tdasilva_ | clayg: yes, now i understand | 21:36 |
clayg | tdasilva_: a test to validate the metadata sent in the PUT doesn't leak into the old version seemed to be missing, maybe I'm off my rocker, this is all from memeory | 21:36 |
*** bitblt has joined #openstack-swift | 21:37 | |
tdasilva_ | clayg: no no, now i understand, thanks! | 21:37 |
clayg | tdasilva_: sweet! no thank YOU! | 21:37 |
tdasilva_ | clayg: for some reason I got caught up in "backend COPY" and I thought you were talking about some backend daemons that I don't know about | 21:37 |
tdasilva_ | clayg: thanks! | 21:38 |
clayg | yeah i'm sure there's something to do with some backend something that's not covered by any test anywhere - but I'm not really sure what to tell you about that... | 21:38 |
tdasilva_ | no, it's ok..makes sense now...I'm in the process of adding more functional tests and I'll make sure to add those tests there too | 21:40 |
clayg | tdasilva_: oh awesome! | 21:40 |
clayg | tdasilva_: yeah I was thinking most of those real socket tests would have looked great/better as functional tests - I think I'd be totally fine dropping them in favor of better functional tests of versioned objects | 21:41 |
tdasilva_ | clayg: you had mentioned that they tested some corner cases, do you remember specifically what they were? | 21:44 |
tdasilva_ | clayg: i've tried going through the tests, but couldn't really point to anything that would be a corner case, so I'm probably missing something | 21:45 |
*** tdasilva has joined #openstack-swift | 21:50 | |
*** tdasilva has quit IRC | 21:58 | |
*** jwang__ has joined #openstack-swift | 22:05 | |
*** rdaly2 has quit IRC | 22:05 | |
*** tellesnobrega_ has joined #openstack-swift | 22:16 | |
*** Masahiro has joined #openstack-swift | 22:21 | |
*** gyee has quit IRC | 22:24 | |
*** Masahiro has quit IRC | 22:26 | |
*** gyee has joined #openstack-swift | 22:27 | |
clayg | tdasilva_: the existing socket tests? they just looked like the poked around the corners - and *they* found the content-type thing - not me really, i just ported them instead of skipping them and they found a bug (although i'll admit it took way too long to tease apart what exactly was failing and why because the stupid readuntil2cr style of requests is a total disaster) | 22:30 |
clayg | tdasilva_: I think they also hit some utf-8 stuff, but all of that seemed to work. | 22:31 |
clayg | tdasilva_: didn't your current change set eat all my ports - so all the existing tests are running and passing? that's about as good as we've got, i don't know off the top of my head any behaviors of versioned objects that are not currently covered by any tests - I just didn't want to loose any coverage we had currently | 22:32 |
clayg | tdasilva_: I guess i thought i saw a gap for preservation of x-object-meta, there's weird combinations of features + object versions that I'm not even sure about the existing behavior - maybe bulk, *lo, expiring + versioned objects | 22:34 |
clayg | notmyname: did the reclaim fix apply pretty cleanly to juno or did you have to un conflict it a bunch? | 22:36 |
notmyname | clayg: cleanly | 22:37 |
*** pberis_ has quit IRC | 22:37 | |
*** pberis has quit IRC | 22:37 | |
notmyname | hmm...but it looks like tests have an issue | 22:38 |
clayg | weeeeeee | 22:38 |
notmyname | oh, maybe it's | 22:38 |
notmyname | or maybe it's not | 22:38 |
*** jwang__ has quit IRC | 22:49 | |
mattoliverau | Yay, I have a concurrent read unit test that only takes 0.703 seconds and doesn't almost crash my desktop.. the secret ended up being using <mock>.read.side_effect = ['custom body',''], so it actually ends and doesn't go into a infinite read() loop. (for those playing along at home) :P | 22:52 |
*** weishao has joined #openstack-swift | 22:53 | |
*** miqui__ has quit IRC | 22:54 | |
*** bkopilov has quit IRC | 22:54 | |
notmyname | mattoliverau: yay | 22:56 |
*** occup4nt has joined #openstack-swift | 22:57 | |
mattoliverau | yay mock-isms | 22:58 |
notmyname | clayg: in the backport, the final part dir isn't getting cleaned up (line 1406 is the backport patch) | 22:59 |
notmyname | clayg: is it assuming the behavior in ctennis's patch (d40cebfe)? | 23:00 |
*** occupant has quit IRC | 23:01 | |
clayg | notmyname: oh... yeah it sure it :\ | 23:01 |
notmyname | clayg: ok, no worries. want me to include that patch or remove the final test in the backport? | 23:01 |
clayg | so that whole run it again and validate the part dir is cleaned up bit can come off, i'm not sure when dfg's remove the suffix dir patch came in? | 23:01 |
clayg | notmyname: can you just modify the test to not have that assert? the part about deleting the db is pretty good - again not sure if we expect the suffix dir to be cleaned up in juno or not? | 23:02 |
notmyname | clayg: actually the patch was from ctennis not dfg | 23:02 |
clayg | no no no, not the remove the *part* dir - the remove the *suffix* dir | 23:02 |
notmyname | oj | 23:03 |
clayg | gawd there's only 4 dirs you'd think you could keep them stright | 23:03 |
*** oomichi has joined #openstack-swift | 23:07 | |
notmyname | ok, bot ctennis's patch and your's proposed as backports to juno. 2 dependent patches | 23:09 |
*** bitblt has quit IRC | 23:18 | |
*** occupant has joined #openstack-swift | 23:18 | |
*** occup4nt has quit IRC | 23:20 | |
*** weishao has quit IRC | 23:33 | |
*** reed has joined #openstack-swift | 23:48 | |
*** mahatic has quit IRC | 23:48 | |
reed | notmyname, I stumbled up this patch, wondering if it needs a blueprint/spec or may not need one https://review.openstack.org/#/c/137659/ | 23:49 |
reed | I am conversating with people at OVH already, I may ask them to add more details, if you think they're needed | 23:49 |
*** geaaru has quit IRC | 23:51 | |
notmyname | reed: need some answers to what clayg asked first. the commit message seems to say it does something that swift has supported for a while, so we want to hear back what they're doing and/or wanting | 23:52 |
*** dmsimard is now known as dmsimard_away | 23:53 | |
reed | ok, I'll suggest them to get on IRC, email and respond to review | 23:53 |
reed | thanks | 23:53 |
notmyname | reed: thank you :-) | 23:53 |
tdasilva_ | clayg: sorry, had to step out for a bit, but I understand everything you said, got enough to keep me going now... | 23:54 |
clayg | tdasilva_: sweet | 23:54 |
*** ahonda has quit IRC | 23:58 | |
*** ahonda has joined #openstack-swift | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!