| stewie925_ | the sudo command failed with that error - we can continue tomorrow? | 00:00 |
|---|---|---|
| notmyname | that's got an easy fix. let me see if I can remember it | 00:00 |
| notmyname | oh. no that was somethign else I was thinking of | 00:01 |
| notmyname | it's this that you're trying to get https://pypi.python.org/pypi/bindep | 00:02 |
| clayg | did you try unplugging your router? | 00:03 |
| notmyname | clayg: that won't work. you gotta plug it back in too! ;-) | 00:03 |
| clayg | :D | 00:03 |
| *** wasmum has joined #openstack-swift | 00:04 | |
| *** hosanai has joined #openstack-swift | 00:05 | |
| *** ChanServ sets mode: +v hosanai | 00:05 | |
| notmyname | ok, I'm taking off. I got to spend time int he swift source tree today instead of writing emails and going to meetings about it. time to call it done while it's all still good :-) | 00:05 |
| *** nadeem has joined #openstack-swift | 00:07 | |
| stewie925_ | talk to you tomorrow, notmyname | 00:07 |
| *** _JZ_ has quit IRC | 00:07 | |
| *** garthb_ has quit IRC | 00:23 | |
| *** nadeem has quit IRC | 00:36 | |
| *** suyash has quit IRC | 00:36 | |
| kota_ | good morning | 00:40 |
| mattoliverau | kota_: morning | 00:50 |
| kota_ | mattoliverau: morning, it looks like a bunch of conversation at IRC yesterday while I'm asleep. | 00:51 |
| mattoliverau | kota_: yeah, looks like the US's long weekend is over ;) | 00:52 |
| kota_ | mattoliverau: gotcha! | 00:52 |
| *** stewie925_ has quit IRC | 00:54 | |
| *** sams-gleb has joined #openstack-swift | 00:55 | |
| *** sams-gleb has quit IRC | 00:59 | |
| *** asettle has joined #openstack-swift | 01:08 | |
| *** asettle has quit IRC | 01:13 | |
| *** klamath has quit IRC | 01:15 | |
| *** klamath has joined #openstack-swift | 01:16 | |
| *** DCWilliams_VA has joined #openstack-swift | 01:20 | |
| *** DCWilliams_VA is now known as Guest99453 | 01:20 | |
| *** Guest99453 is now known as dcwilliams29id | 01:22 | |
| *** dcwilliams29id is now known as dcwilliamsva | 01:23 | |
| *** natarej__ has quit IRC | 01:28 | |
| *** natarej has joined #openstack-swift | 01:29 | |
| *** dcwilliamsva has quit IRC | 01:29 | |
| *** thumpba has joined #openstack-swift | 01:50 | |
| *** klamath has quit IRC | 01:50 | |
| *** klamath has joined #openstack-swift | 01:50 | |
| CarlFK | what is the rackspace version ov authurl='https://objects.dreamhost.com/auth', | 01:53 |
| *** klamath has quit IRC | 01:53 | |
| *** klamath has joined #openstack-swift | 01:54 | |
| *** sams-gleb has joined #openstack-swift | 01:56 | |
| *** sams-gleb has quit IRC | 02:00 | |
| *** tqtran has quit IRC | 02:10 | |
| *** CaioBrentano has joined #openstack-swift | 02:13 | |
| *** hogepodge has quit IRC | 02:13 | |
| *** thumpba has quit IRC | 02:16 | |
| *** hogepodge has joined #openstack-swift | 02:17 | |
| *** CaioBrentano has quit IRC | 02:18 | |
| *** DCWilliams_VA has joined #openstack-swift | 02:21 | |
| *** DCWilliams_VA is now known as Guest35307 | 02:21 | |
| *** hogepodge has quit IRC | 02:21 | |
| redbo | CarlFK: the rackspace authentication endpoint is 'https://identity.api.rackspacecloud.com/v2.0/' | 02:22 |
| CarlFK | redbo: thanks! and I think I just found that here: https://github.com/rackspace/pyrax/blob/master/pyrax/identity/rax_identity.py#L14 AUTH_ENDPOINT = "https://identity.api.rackspacecloud.com/v2.0/" | 02:23 |
| *** dmorita has quit IRC | 02:30 | |
| CarlFK | redbo: I am trying these snippets http://docs.ceph.com/docs/jewel/radosgw/swift/python/ | 02:32 |
| CarlFK | user = 'account_name:username' | 02:32 |
| CarlFK | what's account_name? | 02:32 |
| *** dmorita has joined #openstack-swift | 02:37 | |
| *** dmorita has quit IRC | 02:37 | |
| *** tanee_away is now known as tane | 02:42 | |
| *** tane is now known as tanee | 02:42 | |
| *** gyee has quit IRC | 02:45 | |
| *** sheel has joined #openstack-swift | 02:51 | |
| redbo | CarlFK: You know, I'm not really sure. Actually, the only way I know of to make it work with swiftclient is to get your API key from the rackspace cloud control panel and use v1 auth. | 02:57 |
| redbo | Then it's: c = swiftclient.Connection(user='redbo', key='8175bce8b...', authurl='https://identity.api.rackspacecloud.com/v1.0/') | 02:57 |
| openstackgerrit | Janie Richling proposed openstack/swift: crypto - remove version from derived iv_base https://review.openstack.org/323636 | 02:59 |
| CarlFK | redbo: umm... it worked \o/ | 03:00 |
| CarlFK | which makes no sense given all I did was swap vars for constants... | 03:00 |
| CarlFK | oh... v1 vs v2 | 03:01 |
| *** Guest35307 has quit IRC | 03:04 | |
| redbo | That's not so great because v1 doesn't support multiple regions. But I don't know who to be mad at. Maybe everyone. | 03:05 |
| CarlFK | lol | 03:06 |
| *** DCWilliams_VA has joined #openstack-swift | 03:12 | |
| *** DCWilliams_VA is now known as Guest57552 | 03:13 | |
| *** Guest57552 has quit IRC | 03:23 | |
| *** takashi has joined #openstack-swift | 03:43 | |
| *** sams-gleb has joined #openstack-swift | 03:57 | |
| *** links has joined #openstack-swift | 04:00 | |
| *** sams-gleb has quit IRC | 04:02 | |
| *** jamielennox is now known as jamielennox|away | 04:08 | |
| zaitcev | I thought it was something obvious like zaitcev:zaitcev and then either v1 or v2 | 04:09 |
| *** psachin has joined #openstack-swift | 04:15 | |
| *** david-lyle has quit IRC | 04:21 | |
| *** natarej has quit IRC | 04:25 | |
| *** natarej has joined #openstack-swift | 04:25 | |
| *** natarej has quit IRC | 04:27 | |
| redbo | That would be a reasonable thing to think. | 04:28 |
| *** natarej_ has joined #openstack-swift | 04:29 | |
| *** nadeem has joined #openstack-swift | 04:32 | |
| *** psachin has quit IRC | 04:35 | |
| *** dmorita has joined #openstack-swift | 04:37 | |
| *** dmorita has quit IRC | 04:42 | |
| *** psachin has joined #openstack-swift | 04:47 | |
| redbo | The problem seems to be that swiftclient blows up if you try to use auth v2 and don't specify a tenant. Well, aside than the problems of auth being overly complicated and rackspace auth not using tenants. | 04:48 |
| *** ppai has joined #openstack-swift | 04:52 | |
| *** natarej has joined #openstack-swift | 04:53 | |
| *** natarej_ has quit IRC | 04:54 | |
| *** nadeem has quit IRC | 04:56 | |
| *** tanee is now known as tanee_away | 04:58 | |
| *** tanee_away is now known as tanee | 05:00 | |
| *** tanee is now known as tanee_away | 05:00 | |
| *** tanee_away is now known as tanee | 05:00 | |
| *** tanee is now known as tanee_away | 05:00 | |
| *** SkyRocknRoll has joined #openstack-swift | 05:02 | |
| *** tanee_away is now known as tanee | 05:03 | |
| *** rcernin has quit IRC | 05:04 | |
| CarlFK | lol | 05:09 |
| *** ChubYann has quit IRC | 05:21 | |
| *** asettle has joined #openstack-swift | 05:24 | |
| CarlFK | so this worked: o = conn.put_object('testing','upload.py',contents= hello_file.read(),content_type='text/plain') | 05:27 |
| *** asettle has quit IRC | 05:28 | |
| CarlFK | http://940b8df95a674fe640d8-ea6444471e13d417907e633ec9e9ff11.r89.cf2.rackcdn.com/upload.py | 05:37 |
| CarlFK | how do I get that URL from my code? | 05:38 |
| *** natarej_ has joined #openstack-swift | 05:39 | |
| *** natarej has quit IRC | 05:42 | |
| *** rcernin has joined #openstack-swift | 05:43 | |
| *** sams-gleb has joined #openstack-swift | 05:59 | |
| *** mvk_ has joined #openstack-swift | 06:00 | |
| *** sams-gleb has quit IRC | 06:03 | |
| *** mvk has quit IRC | 06:03 | |
| *** tanee is now known as tanee_away | 06:11 | |
| *** takashi has quit IRC | 06:14 | |
| *** asettle has joined #openstack-swift | 06:16 | |
| *** pietervanw_ has joined #openstack-swift | 06:18 | |
| *** asettle has quit IRC | 06:21 | |
| *** pietervanw_ has quit IRC | 06:22 | |
| *** mmuffinman has quit IRC | 06:29 | |
| *** hogepodge has joined #openstack-swift | 06:31 | |
| *** tanee_away is now known as tanee | 06:35 | |
| *** mmuffinman has joined #openstack-swift | 06:36 | |
| *** jmccarthy1 has joined #openstack-swift | 06:38 | |
| *** mmuffinman has quit IRC | 06:46 | |
| *** sams-gleb has joined #openstack-swift | 06:51 | |
| cschwede | notmyname: alright, thx for the heads up! | 06:56 |
| cschwede | will have a look at -infra | 06:56 |
| cschwede | well, i think the multinode failed due to some error unrelated to my patch… | 06:58 |
| *** klrmn has quit IRC | 07:02 | |
| *** tesseract has joined #openstack-swift | 07:03 | |
| *** rledisez has joined #openstack-swift | 07:10 | |
| *** sams-gleb has quit IRC | 07:22 | |
| *** DCWilliams_VA has joined #openstack-swift | 07:25 | |
| *** DCWilliams_VA is now known as Guest75101 | 07:26 | |
| *** pcaruana has joined #openstack-swift | 07:26 | |
| *** sams-gleb has joined #openstack-swift | 07:27 | |
| *** Guest75101 has quit IRC | 07:30 | |
| *** takashi has joined #openstack-swift | 07:41 | |
| *** daemontool_ has joined #openstack-swift | 07:42 | |
| *** ppai is now known as ppai|afk | 07:47 | |
| *** sams-gleb has quit IRC | 07:48 | |
| *** sams-gleb has joined #openstack-swift | 07:49 | |
| *** elambert_ has joined #openstack-swift | 07:55 | |
| *** tmoreira has quit IRC | 07:58 | |
| elambert_ | tdasilva: what’s the process for submitting PRs to the new liberasurecode github repo? | 08:00 |
| *** tmoreira has joined #openstack-swift | 08:02 | |
| *** joeljwright has joined #openstack-swift | 08:04 | |
| *** ChanServ sets mode: +v joeljwright | 08:04 | |
| *** joeljwright has quit IRC | 08:04 | |
| *** joeljwright has joined #openstack-swift | 08:05 | |
| *** ChanServ sets mode: +v joeljwright | 08:05 | |
| *** mmcardle has joined #openstack-swift | 08:08 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/swift: [WIP] Testing latest u-c https://review.openstack.org/318441 | 08:10 |
| *** jordanP has joined #openstack-swift | 08:27 | |
| *** jordan_ has joined #openstack-swift | 08:27 | |
| *** mmcardle1 has joined #openstack-swift | 08:29 | |
| *** jmccarthy has quit IRC | 08:31 | |
| *** mmcardle has quit IRC | 08:31 | |
| *** jmccarthy has joined #openstack-swift | 08:31 | |
| *** dmk0202 has joined #openstack-swift | 08:38 | |
| *** ppai|afk is now known as ppai | 08:41 | |
| kota_ | elambert_: probably, current liberasurecode PRs will work via openstack review system. | 08:43 |
| kota_ | elambert_: the way to propose patch will be similar with Swift, I found the docs how to propose a patch to upstream here, http://docs.openstack.org/developer/swift/first_contribution_swift.html?highlight=contribute#proposing-changes-to-swift | 08:44 |
| *** kei_yama has quit IRC | 08:45 | |
| kota_ | probably, git clone http://github.com/openstack/liberasurecode.git -> make a commit -> git review will work for you. | 08:45 |
| kota_ | it might be needed to setup your git review and creating account in the gerrit code review system though. | 08:46 |
| onovy | if new liberasure repo is here: https://github.com/openstack/liberasurecode why someone commited to https://bitbucket.org/tsg-/liberasurecode/commits/all? | 08:46 |
| onovy | someone=Eric :] | 08:46 |
| kota_ | onovy: yeah, he is elambert_, I think. | 08:46 |
| onovy | ah, lol :) | 08:47 |
| onovy | so openstack one is official now? can i package it for debian? | 08:47 |
| kota_ | onovy: not sure, I just saw tsg- (pyeclib, liberasurecode maintainer) said https://github.com/openstak/liberasurecode is a current upstream repo. | 08:48 |
| *** dmk0202 has quit IRC | 08:48 | |
| *** daemontool_ has quit IRC | 08:58 | |
| *** daemontool has joined #openstack-swift | 08:59 | |
| *** psachin_ has joined #openstack-swift | 09:00 | |
| *** psachin has quit IRC | 09:01 | |
| *** daemontool has quit IRC | 09:11 | |
| *** daemontool_ has joined #openstack-swift | 09:11 | |
| *** mvk_ has quit IRC | 09:14 | |
| *** asettle has joined #openstack-swift | 09:20 | |
| *** asettle has quit IRC | 09:22 | |
| *** geaaru has joined #openstack-swift | 09:25 | |
| *** mvk has joined #openstack-swift | 09:32 | |
| openstackgerrit | drax proposed openstack/swift: In case the lenght of the data returned by the object service is not equal to required content length then raise exception so that client can close connection from its end. Fixes Bug 1568650 https://review.openstack.org/323756 | 09:33 |
| openstack | bug 1568650 in OpenStack Object Storage (swift) "Connection between client and proxy service does not closes" [High,Confirmed] https://launchpad.net/bugs/1568650 - Assigned to drax (devesh-gupta) | 09:33 |
| *** takashi has quit IRC | 09:35 | |
| openstackgerrit | Merged openstack/swift: add explicit HA info to the deployment guide https://review.openstack.org/323524 | 09:38 |
| *** dmorita has joined #openstack-swift | 09:39 | |
| openstackgerrit | Merged openstack/swift: Don't include holes when reporting how many devices a ring has https://review.openstack.org/321842 | 09:41 |
| *** asettle has joined #openstack-swift | 09:43 | |
| *** dmorita has quit IRC | 09:43 | |
| *** jamielennox|away is now known as jamielennox | 09:44 | |
| *** Jeffrey4l_ has quit IRC | 09:48 | |
| *** asettle has quit IRC | 09:57 | |
| *** daemontool_ has quit IRC | 10:00 | |
| openstackgerrit | Davanum Srinivas (dims) proposed openstack/swift: [WIP] Testing latest u-c https://review.openstack.org/318441 | 10:10 |
| *** elambert_ has quit IRC | 10:22 | |
| *** asettle has joined #openstack-swift | 10:29 | |
| *** wasmum has quit IRC | 10:44 | |
| *** hosanai has quit IRC | 10:44 | |
| openstackgerrit | Alistair Coles proposed openstack/swift: crypto - use random iv when wrapping body key https://review.openstack.org/321451 | 10:45 |
| openstackgerrit | Alistair Coles proposed openstack/swift: crypto - stop using key id to derive GET path keys https://review.openstack.org/318148 | 11:04 |
| *** acoles_ is now known as acoles | 11:05 | |
| *** raildo-afk is now known as raildo | 11:15 | |
| *** ppai has quit IRC | 11:34 | |
| *** links has quit IRC | 11:34 | |
| openstackgerrit | Alistair Coles proposed openstack/swift: crypto - add test to verify ondisk container listing https://review.openstack.org/323355 | 11:37 |
| *** ppai has joined #openstack-swift | 11:41 | |
| *** links has joined #openstack-swift | 11:47 | |
| openstackgerrit | Doron Chen proposed openstack/swift: Added a test for delayed reap. https://review.openstack.org/321041 | 11:55 |
| openstackgerrit | Ondřej Nový proposed openstack/liberasurecode: Changed homepage to new one https://review.openstack.org/323819 | 11:56 |
| tdasilva | onovy: yeah, https://github.com/openstack/liberasurecode is official now | 11:57 |
| tdasilva | i'm working with Tushar to change bitbucket repo to point people to github | 11:57 |
| onovy | tdasilva: thanks! ^^ :) | 11:58 |
| *** wasmum has joined #openstack-swift | 12:05 | |
| *** _JZ_ has joined #openstack-swift | 12:07 | |
| *** _JZ_ has quit IRC | 12:07 | |
| *** _JZ_ has joined #openstack-swift | 12:12 | |
| *** sams-gleb has quit IRC | 12:13 | |
| *** delattec has joined #openstack-swift | 12:15 | |
| *** openstackgerrit has quit IRC | 12:18 | |
| *** ppai has quit IRC | 12:18 | |
| *** cdelatte has quit IRC | 12:18 | |
| *** openstackgerrit has joined #openstack-swift | 12:18 | |
| *** DCWilliams_VA has joined #openstack-swift | 12:19 | |
| *** DCWilliams_VA is now known as Guest29735 | 12:19 | |
| *** Guest29735 has quit IRC | 12:20 | |
| *** pauloewerton has joined #openstack-swift | 12:30 | |
| *** SkyRocknRoll has quit IRC | 12:48 | |
| *** delatte has joined #openstack-swift | 12:52 | |
| *** Jeffrey4l has joined #openstack-swift | 12:53 | |
| *** CaioBrentano has joined #openstack-swift | 12:53 | |
| *** delattec has quit IRC | 12:55 | |
| *** klamath has quit IRC | 13:00 | |
| *** klamath has joined #openstack-swift | 13:00 | |
| *** SkyRocknRoll has joined #openstack-swift | 13:03 | |
| openstackgerrit | Merged openstack/swift: crypto - use random iv when wrapping body key https://review.openstack.org/321451 | 13:05 |
| *** mariusv has quit IRC | 13:10 | |
| *** mariusv has joined #openstack-swift | 13:11 | |
| *** sams-gleb has joined #openstack-swift | 13:13 | |
| *** vinsh_ is now known as Vinsh | 13:15 | |
| onovy | tdasilva: dtto for python-pyeclib repo? | 13:19 |
| openstackgerrit | Felix Cantournet proposed openstack/swift: Fix swift-dispersion in multi-region setups https://review.openstack.org/322820 | 13:34 |
| *** SkyRocknRoll has quit IRC | 13:34 | |
| openstackgerrit | Felix Cantournet proposed openstack/swift: Fix swift-dispersion in multi-region setups https://review.openstack.org/322820 | 13:35 |
| acoles | jrichli: mahatic : is there any reason why crypto_utils.py and crypto.py cannot be a single module? They are both small and both contain code shared between encrypter and decrypter. | 13:36 |
| *** cguadall has joined #openstack-swift | 13:36 | |
| openstackgerrit | Ondřej Nový proposed openstack/pyeclib: Changed liberasure URL to new one https://review.openstack.org/323874 | 13:40 |
| *** dmorita has joined #openstack-swift | 13:40 | |
| openstackgerrit | Alistair Coles proposed openstack/swift: Merge branch 'master' into feature/crypto https://review.openstack.org/323875 | 13:41 |
| mahatic | acoles: we could. having crypto_utils seems a bit more easier to track? other than that, don't really see much difference | 13:42 |
| openstackgerrit | Ondřej Nový proposed openstack/pyeclib: Bump liberasurecode dependency to 1.2.0 version https://review.openstack.org/323877 | 13:43 |
| mahatic | a little more readable, than having one big file with everything in it | 13:43 |
| acoles | mahatic: by "easier to track" do you mean combining to crypto_utils is easier than combining into cryto? | 13:43 |
| acoles | mahatic: oic | 13:44 |
| openstackgerrit | Ondřej Nový proposed openstack/pyeclib: Change liberasure URL to new one https://review.openstack.org/323874 | 13:44 |
| acoles | mahatic: unless it's huge I find it easier not to have to switch between files | 13:44 |
| *** dmorita has quit IRC | 13:44 | |
| *** SkyRocknRoll has joined #openstack-swift | 13:47 | |
| mahatic | acoles: maybe. It's won't be all that huge by combining it now, but maybe for future (if/when). I remember some discussion on IRC wanting to split a file (in tests or some other module) because it's huge. But anyway, for *now* with current code, doesn't seem like a bad idea | 13:48 |
| *** MikeG451 has joined #openstack-swift | 13:48 | |
| mahatic | combining both (crypto_utils and crypto) doesn't seem like a bad idea* | 13:48 |
| *** links has quit IRC | 13:52 | |
| *** klamath has quit IRC | 13:54 | |
| *** klamath has joined #openstack-swift | 13:54 | |
| *** psachin_ has quit IRC | 13:54 | |
| *** klamath has quit IRC | 13:56 | |
| tdasilva | onovy: yes, same for pyeclib | 13:56 |
| *** klamath has joined #openstack-swift | 13:57 | |
| *** ametts has joined #openstack-swift | 13:58 | |
| *** SkyRocknRoll has quit IRC | 13:58 | |
| *** StraubTW has joined #openstack-swift | 14:00 | |
| openstackgerrit | Merged openstack/swift: crypto - add test to verify ondisk container listing https://review.openstack.org/323355 | 14:03 |
| onovy | tdasilva: cool thanks, sent few merge req with nits in liberasurecode + pyeclib | 14:11 |
| *** thumpba has joined #openstack-swift | 14:11 | |
| *** psachin_ has joined #openstack-swift | 14:14 | |
| *** zul has quit IRC | 14:18 | |
| notmyname | good morning | 14:21 |
| *** zul has joined #openstack-swift | 14:21 | |
| *** Raymii has quit IRC | 14:21 | |
| tdasilva | onovy: cool, just saw that | 14:22 |
| *** thumpba has quit IRC | 14:25 | |
| *** d0ugal has quit IRC | 14:25 | |
| *** diogogmt has joined #openstack-swift | 14:27 | |
| *** ametts has quit IRC | 14:28 | |
| openstackgerrit | Christian Schwede proposed openstack/swift: Refactor locale tests and unicode issue https://review.openstack.org/323946 | 14:30 |
| openstackgerrit | Christian Schwede proposed openstack/swift: Refactor locale tests and unicode issue https://review.openstack.org/323950 | 14:32 |
| openstackgerrit | Alistair Coles proposed openstack/swift: crypto - purge crypto sysmeta from responses https://review.openstack.org/320579 | 14:33 |
| *** d0ugal has joined #openstack-swift | 14:35 | |
| *** thumpba has joined #openstack-swift | 14:36 | |
| jrichli | acoles mahatic: I had originally thought of crypto.py as holding the code that is tied to a cryptographic implementation. Then the utils was used for methods related to crypto code, but still crypto alg agnostic. | 14:39 |
| jrichli | acoles mahatic: but, it isn't like we have a plugin option for crypto at this point. so, I am game for whatever. | 14:40 |
| *** ametts has joined #openstack-swift | 14:42 | |
| *** StraubTW has quit IRC | 14:46 | |
| openstackgerrit | Mohit Motiani proposed openstack/swift: Add region in search-value of swift-ring-builder manpage https://review.openstack.org/323580 | 14:47 |
| *** d0ugal has quit IRC | 14:48 | |
| *** diogogmt has quit IRC | 14:48 | |
| *** d0ugal has joined #openstack-swift | 14:48 | |
| *** d0ugal has joined #openstack-swift | 14:49 | |
| *** elambert_ has joined #openstack-swift | 14:50 | |
| *** arch-nemesis has joined #openstack-swift | 14:53 | |
| elambert_ | kota_: thanks for the pointers :-) | 15:00 |
| *** diogogmt has joined #openstack-swift | 15:00 | |
| *** diogogmt has quit IRC | 15:01 | |
| *** pcaruana has quit IRC | 15:04 | |
| *** asettle has quit IRC | 15:06 | |
| openstackgerrit | Mohit Motiani proposed openstack/swift: Adds region as a search-value in manpage https://review.openstack.org/323580 | 15:08 |
| *** rcernin has quit IRC | 15:09 | |
| *** d0ugal has quit IRC | 15:10 | |
| *** StraubTW has joined #openstack-swift | 15:11 | |
| openstackgerrit | Nandini Tata proposed openstack/python-swiftclient: Fixed a typo in SwiftService API doc https://review.openstack.org/323975 | 15:12 |
| *** d0ugal has joined #openstack-swift | 15:12 | |
| *** StraubTW has quit IRC | 15:13 | |
| *** asettle has joined #openstack-swift | 15:15 | |
| *** Raymii has joined #openstack-swift | 15:15 | |
| openstackgerrit | Mohit Motiani proposed openstack/swift: Add description of server options (all, main, rest) for swift-init help https://review.openstack.org/316448 | 15:16 |
| *** tesseract has quit IRC | 15:18 | |
| *** ouchkernel has joined #openstack-swift | 15:20 | |
| acoles | jrichli: ack. good point re leading '/'. keep it there. | 15:21 |
| openstackgerrit | Alistair Coles proposed openstack/swift: crypto combine crypto.py with crypto_utils.py https://review.openstack.org/323978 | 15:23 |
| acoles | jrichli: mahatic ^^ | 15:24 |
| *** nadeem has joined #openstack-swift | 15:24 | |
| *** klrmn has joined #openstack-swift | 15:25 | |
| *** permalac has joined #openstack-swift | 15:32 | |
| *** diogogmt has joined #openstack-swift | 15:37 | |
| *** mmuffinman has joined #openstack-swift | 15:38 | |
| *** nadeem has quit IRC | 15:56 | |
| *** sams-gleb has quit IRC | 16:02 | |
| *** geaaru has quit IRC | 16:06 | |
| *** gyee has joined #openstack-swift | 16:09 | |
| openstackgerrit | Janie Richling proposed openstack/swift: crypto - remove version from derived iv_base https://review.openstack.org/323636 | 16:11 |
| *** dmorita has joined #openstack-swift | 16:11 | |
| *** StraubTW has joined #openstack-swift | 16:13 | |
| *** StraubTW has quit IRC | 16:18 | |
| notmyname | joeljwright: did you notice that http://docs.openstack.org/developer/swift/ has the versions on the side now? | 16:19 |
| joeljwright | notmyname: did we abandon the client docs patch in favour of the oslosphinx patch | 16:20 |
| notmyname | I just did | 16:20 |
| joeljwright | notmyname: yes I did, and I got jealous that the swiftclient didn't! | 16:20 |
| notmyname | so I think next time the client docs are rebuilt we'll have it there too | 16:20 |
| notmyname | so we just need your patch to land :-) | 16:20 |
| joeljwright | excellent, I need to reply to that review with a link to the oslosphinx patch | 16:21 |
| *** klrmn has quit IRC | 16:21 | |
| notmyname | I can if you're in the middle of something | 16:22 |
| notmyname | I've got all the links open currently | 16:22 |
| joeljwright | I'm doing it now | 16:22 |
| notmyname | great | 16:22 |
| joeljwright | just wanted to speak to you before I replied | 16:22 |
| notmyname | timburke: around today? | 16:24 |
| *** SkyRocknRoll has joined #openstack-swift | 16:26 | |
| *** elambert_ has quit IRC | 16:26 | |
| notmyname | joeljwright: the sad part is those links don't work :-( | 16:30 |
| joeljwright | damn! | 16:30 |
| joeljwright | ah, yeah! | 16:30 |
| joeljwright | they don't link within the project | 16:31 |
| notmyname | right | 16:31 |
| notmyname | looking at the code to see what to fix | 16:31 |
| notmyname | href="/{{ver}}" | 16:31 |
| notmyname | bah | 16:31 |
| joeljwright | I was doing the same thing :) | 16:32 |
| notmyname | needs pathto(master_doc) I think | 16:32 |
| joeljwright | yeah | 16:33 |
| joeljwright | pathto(master_doc)/ver | 16:33 |
| joeljwright | or something similar | 16:33 |
| notmyname | {{ pathto(master_doc) }}/{{ver}} | 16:33 |
| notmyname | right? | 16:33 |
| notmyname | {{ pathto(master_doc) }}/{{ ver }} | 16:33 |
| notmyname | for spaces consistency | 16:33 |
| joeljwright | the latter looks more promising | 16:34 |
| tdasilva | acoles, jrichli: just a quick reminder, are there specific crypto functional tests? | 16:34 |
| *** nadeem has joined #openstack-swift | 16:34 | |
| tdasilva | or mostly unit, probe tests? | 16:34 |
| joeljwright | although the spacing isn't exactly consistent in the 10 or so lines either side | 16:34 |
| *** nadeem has quit IRC | 16:34 | |
| notmyname | joeljwright: no, that won't work either | 16:34 |
| notmyname | seems that includes "index.html" | 16:35 |
| jrichli | tdasilva: no functests have been added, but we do have a test_encrypter_decrypter that tests the "round_trip" cases. | 16:35 |
| *** nadeem has joined #openstack-swift | 16:35 | |
| *** SkyRocknRoll has quit IRC | 16:35 | |
| notmyname | joeljwright: but master_doc = 'index' so I wonder how pathto() is defined | 16:36 |
| *** jordan_ has quit IRC | 16:36 | |
| tdasilva | jrichli: ok, thanks | 16:36 |
| acoles | tdasilva: we add to functional tests but the additions are not crypto-specific | 16:37 |
| tdasilva | acoles, jrichli: ok, was just thinking it if would make sense to add some func tests to patch 320579 | 16:38 |
| patchbot | tdasilva: https://review.openstack.org/#/c/320579/ - swift (feature/crypto) - crypto - purge crypto sysmeta from responses | 16:38 |
| tdasilva | thinking of the copy case, but seems like that is getting covered with unit tests??? | 16:38 |
| tdasilva | and it's not really crypto specific | 16:38 |
| acoles | tdasilva: the gatekeeper would have purged those sysmeta headers anyway | 16:39 |
| tdasilva | acoles: right | 16:39 |
| acoles | tdasilva: so we're purging them mainly to deal with copy from encrypted to unencrypted which a func test could not cover | 16:40 |
| acoles | tdasilva: but the tests we now have in test_encrypter_decrypter are pretty good for exploring those scenarions | 16:40 |
| acoles | scenarios* | 16:40 |
| * tdasilva goes looking for test_encrypter_decrypter | 16:40 | |
| acoles | tdasilva: https://review.openstack.org/#/c/320579/6/test/unit/common/middleware/test_encrypter_decrypter.py@431 | 16:41 |
| patchbot | acoles: patch 320579 - swift (feature/crypto) - crypto - purge crypto sysmeta from responses | 16:41 |
| notmyname | joeljwright: actually I think maybe we can just drop the initial / | 16:42 |
| joeljwright | notmyname: I thought that, but if the index.html is there (or #) then the link will break | 16:42 |
| tdasilva | acoles: thanks! | 16:42 |
| notmyname | joeljwright: no, it will be relative to the current directory and will replace the index.html | 16:43 |
| notmyname | joeljwright: which "works" because the whole site is only one level deep. no nested directories | 16:43 |
| notmyname | joeljwright: except, of course, the version links. so that would probably end up with 2.7.0/2.6.0/2.5.0 etc | 16:44 |
| joeljwright | notmyname: it's worth a try! | 16:44 |
| notmyname | also, it seems there needs to be a little more filtering in the tags too | 16:44 |
| notmyname | eg filtering out 'rc' | 16:45 |
| notmyname | also, probably should make the first one a link to "current" or something | 16:45 |
| joeljwright | that would certainly make more sense | 16:46 |
| notmyname | ok. I think that's done. | 16:47 |
| notmyname | now to figure out the relative links... | 16:47 |
| joeljwright | I'm hunting in the sphinx docs | 16:47 |
| *** wasmum has quit IRC | 16:49 | |
| notmyname | I don't see anywhere in conf.py (in swift or oslosphinx) where the root is set | 16:49 |
| notmyname | I'm afraid each one may be generated as a local whole-site thing and then simply uploaded to a subdirectory on the server | 16:52 |
| joeljwright | that would be… frustrating | 16:52 |
| *** rledisez has quit IRC | 16:53 | |
| joeljwright | they do use "/developer/{{ project }}" at the end of the file for page tracking... | 16:54 |
| joeljwright | but is "/developer/{{ project }}" always where the docs go for every project? | 16:55 |
| notmyname | oh, good call | 16:55 |
| *** thumpba has quit IRC | 16:56 | |
| notmyname | better than what I was thinking (some python in the template like prefix = {{pathto(master_doc)[:len(pathto(master_doc)]}}, assuming that I could even do the template expansion in the python segments | 16:56 |
| joeljwright | :D | 16:57 |
| notmyname | well, those aren't the right slice offests, but you get the point | 16:57 |
| joeljwright | yeah, I thought that probably wouldn't pass a code review :) | 16:57 |
| *** gyee has quit IRC | 16:58 | |
| *** elambert_ has joined #openstack-swift | 16:59 | |
| notmyname | patch 324024 | 16:59 |
| patchbot | notmyname: https://review.openstack.org/#/c/324024/ - oslosphinx - fix other versions sidebar links | 16:59 |
| *** thumpba has joined #openstack-swift | 17:01 | |
| *** links has joined #openstack-swift | 17:01 | |
| notmyname | joeljwright: seem reasonable to you? | 17:02 |
| *** sams-gleb has joined #openstack-swift | 17:03 | |
| joeljwright | looks good to me | 17:03 |
| joeljwright | will add my +1 | 17:03 |
| joeljwright | notmyname: just one thing - is 'Other Versions' now correct? | 17:04 |
| joeljwright | Seems like it's just 'Versions' now | 17:05 |
| joeljwright | and if I was being really picky, I'd want to know that 'current' means 'current HEAD' | 17:05 |
| joeljwright | :) | 17:05 |
| *** sams-gleb has quit IRC | 17:07 | |
| notmyname | no, current doesn't always mean HEAD | 17:10 |
| joeljwright | ok | 17:10 |
| notmyname | for libraries it could be last released one | 17:10 |
| notmyname | but for services it's current HEAD of master | 17:10 |
| joeljwright | yeah, good point | 17:10 |
| notmyname | just not for swift/swiftclient ;-) | 17:10 |
| joeljwright | ok, I'll just +1 as it is then :) | 17:11 |
| joeljwright | done | 17:12 |
| *** thumpba has quit IRC | 17:14 | |
| *** StraubTW has joined #openstack-swift | 17:16 | |
| *** klrmn has joined #openstack-swift | 17:17 | |
| *** asettle has quit IRC | 17:17 | |
| *** asettle has joined #openstack-swift | 17:18 | |
| *** ChubYann has joined #openstack-swift | 17:19 | |
| *** StraubTW has quit IRC | 17:21 | |
| notmyname | FYI https://bugs.launchpad.net/openstack-manuals/+bug/1588002 | 17:25 |
| openstack | Launchpad bug 1588002 in openstack-manuals "Swift missing from OpenStack High Availability Guide" [Undecided,New] | 17:25 |
| notmyname | timburke: ^ | 17:25 |
| *** joeljwright has quit IRC | 17:33 | |
| *** thumpba has joined #openstack-swift | 17:34 | |
| *** SkyRocknRoll has joined #openstack-swift | 17:34 | |
| *** links has quit IRC | 17:36 | |
| *** wasmum has joined #openstack-swift | 17:42 | |
| *** mvk has quit IRC | 17:48 | |
| *** tqtran has joined #openstack-swift | 17:48 | |
| *** arch-nemesis has quit IRC | 17:50 | |
| *** suyash has joined #openstack-swift | 17:51 | |
| *** arch-nemesis has joined #openstack-swift | 17:51 | |
| *** psachin_ has quit IRC | 17:54 | |
| *** jmccarthy1 has quit IRC | 17:55 | |
| *** thumpba has quit IRC | 17:55 | |
| *** mmcardle1 has quit IRC | 17:56 | |
| *** mmcardle has joined #openstack-swift | 18:00 | |
| notmyname | onovy: around? | 18:00 |
| notmyname | onovy: for when you read scollback.. | 18:01 |
| notmyname | onovy: bad news, I'm afraid. I hadn't forgotten about your request to backport https://review.openstack.org/#/c/270233/ to juno, even though juno is EOL | 18:02 |
| patchbot | notmyname: patch 270233 - swift - Fix memory/socket leak in proxy on truncated SLO/D... (MERGED) | 18:02 |
| notmyname | onovy: so I was just looking in to that, and, well, it's not going to happen. there's quit a bit of changes that start reaching throughout the system | 18:02 |
| *** elambert_ has quit IRC | 18:02 | |
| notmyname | onovy: it's not at all a small isolated fix, so it's not something I want to pursue further | 18:03 |
| *** elambert_ has joined #openstack-swift | 18:04 | |
| *** thumpba has joined #openstack-swift | 18:15 | |
| zaitcev | oi | 18:15 |
| zaitcev | I just noticed that we don't have an object-expirer.conf-sample | 18:15 |
| timburke | good morning | 18:19 |
| zaitcev | Yet I have a seemingly functional expirer in my cluster with a configuration that came from nowhere. | 18:22 |
| *** mmcardle has quit IRC | 18:26 | |
| *** mvk has joined #openstack-swift | 18:39 | |
| *** Jeffrey4l has quit IRC | 18:42 | |
| *** Jeffrey4l has joined #openstack-swift | 18:43 | |
| *** mvk_ has joined #openstack-swift | 18:44 | |
| acoles | tdasilva: your review comment got me trying to figure how a COPY with Range succeeds on master given that the etag for the PUT will be wrong...and it turns out that (perhaps fortuitously) this line has no effect https://github.com/openstack/swift/blob/eda1b330f3cf7bf9dd643854d4a221b6f024884d/swift/common/middleware/copy.py#L477-L477 | 18:46 |
| timburke | acoles: i'd been wondering about that... but more in the context of flattening large objects | 18:47 |
| *** mvk has quit IRC | 18:48 | |
| acoles | tdasilva: so your comment here is a valid concern, that we need to be sure not to send an etag in an override header for a ranged COPY https://review.openstack.org/#/c/320579/6/test/unit/common/middleware/test_copy.py@610 | 18:48 |
| patchbot | acoles: patch 320579 - swift (feature/crypto) - crypto - purge crypto sysmeta from responses | 18:48 |
| *** mvk has joined #openstack-swift | 18:48 | |
| *** mvk_ has quit IRC | 18:48 | |
| acoles | timburke: I am sure I've seen this req.etag thing done before and we fixed it ?!? | 18:49 |
| openstackgerrit | Merged openstack/python-swiftclient: Migrated contributing file from .md to .rst https://review.openstack.org/316758 | 18:50 |
| acoles | timburke: so tdasilva's review comment ^^ takes us back to whether it is safe to "copy" the override sysmeta headers or whether we should purge them on a copy? Seems like the ranged copy is a case when the override etag value will be wrong for the PUT path | 18:50 |
| acoles | timburke: but if that is the case it also suggests we don't have a functional test that does a copy with Range header :/ cos it would fail | 18:51 |
| *** mvk_ has joined #openstack-swift | 18:52 | |
| * acoles back later for meeting | 18:53 | |
| *** acoles is now known as acoles_ | 18:54 | |
| *** ChubYann has quit IRC | 18:56 | |
| *** mvk has quit IRC | 18:56 | |
| notmyname | jrichli: the todo list on https://trello.com/b/63l5zQhq/swift-encryption is really short | 18:59 |
| *** thumpba has quit IRC | 18:59 | |
| *** tqtran has quit IRC | 19:00 | |
| *** thumpba has joined #openstack-swift | 19:00 | |
| *** tqtran has joined #openstack-swift | 19:00 | |
| jrichli | notmyname : yep! I hope clayg is getting prepared for his 2 weeks of review he spoke to me about in Bristol ;-) | 19:00 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/python-swiftclient: Updated from global requirements https://review.openstack.org/89250 | 19:06 |
| *** ChubYann has joined #openstack-swift | 19:09 | |
| cschwede | zaitcev: i see an object-expirer.conf-sample example on master: https://github.com/openstack/swift/blob/master/etc/object-expirer.conf-sample ? | 19:13 |
| *** SkyRocknRoll has quit IRC | 19:14 | |
| zaitcev | cschwede: very mysterious | 19:14 |
| zaitcev | OMG | 19:15 |
| zaitcev | It was a detached head. I researched something... | 19:15 |
| zaitcev | Oh god, never again. | 19:15 |
| clayg | jrichli: i recall nothing of the sort | 19:26 |
| openstackgerrit | Mohit Motiani proposed openstack/swift: Add region in ring structure & deployment guide https://review.openstack.org/320146 | 19:29 |
| *** permalac has quit IRC | 19:30 | |
| *** Jeffrey4l_ has joined #openstack-swift | 19:31 | |
| *** Jeffrey4l has quit IRC | 19:34 | |
| *** djc_ has joined #openstack-swift | 19:37 | |
| djc_ | Is there an optimal number of objects per container or is it unlimited? | 19:37 |
| klamath | ive heard 500k is a sweet spot | 19:38 |
| djc_ | klamath: does latency increase after the 500k object? | 19:39 |
| klamath | performance starts to suffer, yes | 19:39 |
| klamath | ive found 1.7m swift really starts getting bent out of shape in a very small lab i tested on | 19:39 |
| jrichli | clayg: https://media.giphy.com/media/FAX99ve3Egvok/giphy.gif | 19:40 |
| klamath | sharding is always best | 19:40 |
| djc_ | klamath: is that because the filesystem on the object server becomes to large? | 19:41 |
| klamath | not a file system, but a sqlite DB that account/container uses | 19:41 |
| djc_ | klamath: can you provide more details on sharding? | 19:42 |
| klamath | its a general idea about multiplexing your writes into multiple containers at once | 19:43 |
| tdasilva | acoles_: just saw your msgs now, sounds like we missing some func tests in copy | 19:43 |
| klamath | if you can do multiple accounts + containers the better | 19:43 |
| klamath | djc_: http://www.gossamer-threads.com/lists/openstack/dev/33820 | 19:43 |
| klamath | djc_: if you run your own swift cluster offloading account/container to SSD or even raid0 SSD would help speed up write operations if your going to bin pack containers | 19:45 |
| *** sheel has quit IRC | 19:45 | |
| djc_ | klamath: I believe our account/container is on hybrd | 19:48 |
| klamath | ouch | 19:49 |
| djc_ | klamath: better than hdd (spinning disk) though. :) | 19:49 |
| klamath | i dont think the hybrid part will come into play at all | 19:50 |
| klamath | if it does it might just cause the queue to rise with random seeks as it tries to flush stuff to disk | 19:50 |
| klamath | as always test | 19:50 |
| openstackgerrit | Mohit Motiani proposed openstack/swift: Add region in ring structure & deployment guide https://review.openstack.org/320146 | 19:53 |
| clayg | torgomatic: it is *so* hard to unravl the iterators on GET these days :'( | 19:54 |
| clayg | torgomatic: lp bug #1568650 is legit tho | 19:54 |
| openstack | Launchpad bug 1568650 in OpenStack Object Storage (swift) "Connection between client and proxy service does not closes" [High,In progress] https://launchpad.net/bugs/1568650 - Assigned to drax (devesh-gupta) | 19:54 |
| clayg | torgomatic: I can see now that when a backend connection just starts giving up chunks (returns '') that the proxy doesn't really realize that it hasn't given the client all the bytes it promised | 19:55 |
| djc_ | klamath: thanks | 19:56 |
| clayg | torgomatic: the code all seems to assume that the backend would give us a ChunkReadTimeout before a '' - but i think if the client is reading slow enough the kernel has enough time to realize the connection is closed and shut it down so there's a clear empty read (signaling end of file) rather than a timeout | 19:56 |
| clayg | torgomatic: I'm pretty sure those part_iter_from_response iteration knows how big the response is going to be - so my thought is to just make the part_file that gets returned known how many bytes it should be able to read and return raise an error rather than return an empty chunk if didn't get all the expected bytes | 19:58 |
| *** stewie925 has joined #openstack-swift | 19:58 | |
| clayg | torgomatic: but I don't know which of the variable names (I guess "length"?) would be the right guy to watch bytes from? | 19:59 |
| openstackgerrit | Eric Lambert proposed openstack/liberasurecode: Fix clang compile time error https://review.openstack.org/324083 | 20:03 |
| notmyname | elambert_: so you're back in the swift world again? | 20:05 |
| elambert_ | notmyname: yeah but not as part of my day job | 20:05 |
| *** david-lyle has joined #openstack-swift | 20:05 | |
| notmyname | elambert_: ah, ok. swift-as-hobby? ;-) | 20:06 |
| openstackgerrit | Merged openstack/python-swiftclient: Fixed a typo in SwiftService API doc https://review.openstack.org/323975 | 20:06 |
| elambert_ | :—) | 20:06 |
| stewie925 | hello notmyname :) | 20:06 |
| * tdasilva is driving home, will be back for meeting | 20:09 | |
| notmyname | tdasilva: watch out for all that crazy traffic in westford | 20:09 |
| clayg | notmyname: i don't get it | 20:10 |
| notmyname | clayg: he said one of the greatest thigns about living there is his commute is like 5 minutes on back-country roads | 20:10 |
| clayg | notmyname: oh i see - so your comment was *sarcastic* - i get it now - funny | 20:11 |
| clayg | timburke: gd, what happened here!? patch 280200 | 20:30 |
| patchbot | clayg: https://review.openstack.org/#/c/280200/ - python-swiftclient - Add copy object method | 20:30 |
| *** NM has joined #openstack-swift | 20:31 | |
| openstackgerrit | Merged openstack/swift: crypto - remove version from derived iv_base https://review.openstack.org/323636 | 20:36 |
| openstackgerrit | Nelson Marcos de Almeida proposed openstack/swift: Adding sorting_method to admin_guide https://review.openstack.org/324097 | 20:38 |
| *** cutforth has joined #openstack-swift | 20:43 | |
| *** david-lyle has quit IRC | 20:51 | |
| clayg | is the swift-meeting in like 8 mins? | 20:52 |
| notmyname | clayg: yes | 20:52 |
| clayg | k, might be late - ya'll have fun | 20:52 |
| *** thumpba has quit IRC | 20:54 | |
| torgomatic | clayg: oh, I see what you're saying... the object server is running along quickly, so it notices the socket write timed out and kills the connection | 20:56 |
| torgomatic | so then the proxy just gets empty reads on its socket... yeah, that'll goof things up | 20:56 |
| torgomatic | although wouldn't the proxy's iterator then just be finished, and all the stuff goes out of scope and gets freed? I don't see why this would cause any sort of socket leak. | 20:58 |
| notmyname | just about meeting time (#openstack-meeting) | 20:59 |
| *** acoles_ is now known as acoles | 20:59 | |
| mattoliverau | Morning | 20:59 |
| acoles | mattoliverau: o/ | 20:59 |
| *** joeljwright has joined #openstack-swift | 21:00 | |
| *** ChanServ sets mode: +v joeljwright | 21:00 | |
| *** david-lyle has joined #openstack-swift | 21:01 | |
| *** raildo is now known as raildo-afk | 21:03 | |
| *** m_kazuhiro has joined #openstack-swift | 21:07 | |
| clayg | torgomatic: right no leak, but the client doesn't realize the proxy isn't going to send more bytes | 21:09 |
| torgomatic | clayg: oh, right. yeah, that's not good. | 21:09 |
| clayg | torgomatic: so a well behaved client will hang on read - while the proxy on the otherhand is waiting on the next pipelined request | 21:09 |
| openstackgerrit | Merged openstack/swift: Merge branch 'master' into feature/crypto https://review.openstack.org/323875 | 21:11 |
| *** CarlFK has quit IRC | 21:17 | |
| *** pauloewerton has quit IRC | 21:32 | |
| *** bill_az has joined #openstack-swift | 21:34 | |
| *** NM has quit IRC | 21:34 | |
| *** djc_ has quit IRC | 21:36 | |
| clayg | timburke: so the regular PUT Object - Copy docs have less to say about Range headers -> http://docs.aws.amazon.com/AmazonS3/latest/API/RESTObjectCOPY.html | 21:37 |
| timburke | clayg: they don't support it. only upload part does. no idea why | 21:37 |
| clayg | does x-amz-copy-source-range *ever* support multi-range? | 21:38 |
| clayg | i guess it looks like it always has to be bytes=x-y - i can't even tell if they support "bytes=-y" or "bytes=x-" ??? | 21:39 |
| clayg | timburke: GET requests seem to respect the full range specification? | 21:39 |
| timburke | clayg: no, they only ever seem to support a single range. i don't remember how well they support the open-ended ranges. and yeah, GET should be able to do all the expected stuff | 21:40 |
| clayg | timburke: could the use-case be covered by - create a SLOrange - then COPY the SLOrange to a new object?! | 21:43 |
| timburke | clayg: which use-case? probably? but it adds even more steps to the operation | 21:45 |
| clayg | but... on the plus side - well defined and we don't have to hi-jack any existing HTTP/WebDav symantics or invent new headers to remain un-ambiguous! | 21:49 |
| notmyname | ranged copy question added to next week's agenda | 22:00 |
| acoles | good night | 22:00 |
| *** CaioBrentano has quit IRC | 22:00 | |
| *** CaioBrentano has joined #openstack-swift | 22:01 | |
| *** acoles is now known as acoles_ | 22:01 | |
| *** david-lyle has quit IRC | 22:01 | |
| torgomatic | huh, I wonder if the support for x-amz-copy-source-range on S3's Upload Part request is their way of matching SLO + ranges | 22:01 |
| clayg | notmyname: you should tell tom to add this to a wiki page somewhere -> https://hbr.org/resources/images/article_assets/2015/11/R1512E_MEYER_COUNTERPART.png | 22:02 |
| clayg | very good stuff | 22:02 |
| clayg | torgomatic: totally! | 22:02 |
| notmyname | clayg: the crazy thing is that we have people from ALL FOUR quadrants! | 22:02 |
| notmyname | normally in the same room at the same time | 22:02 |
| notmyname | it's amazing anything gets done at all ;-) | 22:03 |
| clayg | timburke: I've often considered if a SLO request really need be stored in swift? DLO too? What if you could just give swift a body defining the manifest and it responded back with the bytes!? May lead way for a COPY request that's more POST like - you give me the read plan and I will a) give you the bytes b) store the bytes in the location of your choosing | 22:04 |
| clayg | notmyname: +1 it's impossible to explain | 22:04 |
| notmyname | clayg: and FWIW I don't think you fit the "right in the middle" model that your US-ness would indicate ;-) | 22:05 |
| clayg | notmyname: I want to totally cross reference comments in my reviews for upgraders downgraders now - I'm freaking out how many times kota_ has told me "I'm not sure I fully agree here" and I mis-read that like I would have wrote it instead of translating to something I would have said like "this is completly broken" | 22:05 |
| timburke | clayg: that's an interesting idea... | 22:05 |
| *** sams-gleb has joined #openstack-swift | 22:08 | |
| *** sams-gleb has quit IRC | 22:12 | |
| openstackgerrit | Merged openstack/swift: Send correct size in POST async update for EC object https://review.openstack.org/317475 | 22:15 |
| notmyname | yay | 22:15 |
| notmyname | actually, I think that's the last thing I was waiting for before doing a release | 22:16 |
| notmyname | I should do the authors/changelog updates | 22:16 |
| *** david-lyle has joined #openstack-swift | 22:16 | |
| *** StraubTW has joined #openstack-swift | 22:19 | |
| *** m_kazuhiro has quit IRC | 22:22 | |
| *** StraubTW has quit IRC | 22:23 | |
| *** ametts has quit IRC | 22:30 | |
| *** darrenc is now known as darrenc_afk | 22:32 | |
| *** joeljwright has quit IRC | 22:37 | |
| *** darrenc_afk is now known as darrenc | 22:39 | |
| *** MikeG451 has quit IRC | 22:40 | |
| stewie925 | hi notmyname | 22:41 |
| notmyname | hi | 22:46 |
| *** cutforth has quit IRC | 22:47 | |
| stewie925 | hi I'm trying to continue working on that https://review.openstack.org/#/c/319988/ | 22:54 |
| patchbot | stewie925: patch 319988 - swift - get_part_nodes should raise error on invalid part | 22:54 |
| stewie925 | to test that ValueError is raised if the 'part' value is not within range | 22:56 |
| *** arch-nemesis has quit IRC | 22:57 | |
| stewie925 | two lines of code #1 part = -1 #2 self.assertRaises(ValueError,get_part_nodes(part)) | 22:57 |
| notmyname | ok | 22:58 |
| stewie925 | oops - got pep8 error , fixing it | 23:03 |
| *** tqtran has quit IRC | 23:03 | |
| timburke | stewie925: i think you want self.assertRaises(ValueError,get_part_nodes, part) - assertRaises expects the first arg to be an exception, the second arg to be a callable, and passes everything else to the callable; as you've got it now, it will call get_part_nodes (and presumably raise a ValueError) before ever entering assertRaises | 23:06 |
| stewie925 | oh hi timburke, thank you! | 23:07 |
| *** NM has joined #openstack-swift | 23:13 | |
| stewie925 | I have run tox -e py27 tests for my changes... git diff in http://paste.openstack.org/show/507053/ and test results (with error) in http://paste.openstack.org/show/507052/ | 23:26 |
| *** fbo has quit IRC | 23:26 | |
| stewie925 | hi timburke and notmyname - your inputs are appreciated :) | 23:32 |
| *** diogogmt has quit IRC | 23:33 | |
| openstackgerrit | Samuel Merritt proposed openstack/swift: Support binding to Unix domain sockets. https://review.openstack.org/324136 | 23:33 |
| *** kei_yama has joined #openstack-swift | 23:34 | |
| notmyname | torgomatic: interesting. I figured that would be a bigger patch | 23:34 |
| torgomatic | notmyname: the hard work's all in eventlet | 23:35 |
| torgomatic | also, you can't test it yet without my fork of eventlet, so let me go indicate that | 23:35 |
| clayg | stewie925: it's sort of a hozer - in obj.test_reconstructor the policies get some fake rings assigned to them and the replica2part2dev tables are sort of bullshit | 23:38 |
| stewie925 | hello clayg | 23:38 |
| clayg | it says the intended_part_shift is 30, but then it only assins parts 0, 1, 2 - so 3 blows up | 23:38 |
| *** nadeem has quit IRC | 23:39 | |
| stewie925 | oh I see | 23:40 |
| clayg | stewie925: try this -> https://gist.github.com/clayg/500e4330362151a2fc239da4045ba9b2 | 23:44 |
| clayg | stewie925: i'm not quite sure how the test_delete_policy_override_params one is related | 23:45 |
| stewie925 | hi thank you clayg, let me apply this patch and run the test again | 23:46 |
| clayg | stewie925: I think that second test is acctually showing you something worth thinking about - the code int he replicator that uncovered this error is sort of annoying - I think if you raised a really specific InvalidPartitionError (not inherit from ValueError) it would break out of the code path better and you could see why the dir for part "99" isn't getting cleaned up | 23:51 |
| clayg | stewie925: but really the fix is in that block in the replicator that blowing up - it should handle the exception from get_part_nodes a little better - but i'm not even sure what the desired behavior would be? | 23:52 |
| clayg | in the one case i know of where someone found this case in a real deployment it was a lab environment where they had some left over data on their object devices from a previous ring - and it was cirtainly correct to remove them - but I'm not sure that's generally a defnesible behavior - maybe quarantine? | 23:53 |
| clayg | definately want to log a more specific error | 23:53 |
| clayg | either way I think it throws weight towards maybe *not* subclassing ValueError - although i'm not sure what that code path *thought* it was handling? | 23:54 |
| *** _JZ_ has quit IRC | 23:56 | |
| *** _JZ_ has joined #openstack-swift | 23:57 | |
| stewie925 | oh I see , yikes | 23:57 |
| *** _JZ_ has quit IRC | 23:58 | |
| *** david-lyle has quit IRC | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
