| *** sanchitmalhotra has joined #openstack-swift | 00:06 | |
| *** sanchitmalhotra1 has quit IRC | 00:08 | |
| *** changbl has joined #openstack-swift | 00:17 | |
| *** changbl has quit IRC | 00:17 | |
| *** changbl has joined #openstack-swift | 00:17 | |
| *** changbl has joined #openstack-swift | 00:18 | |
| *** changbl_ has joined #openstack-swift | 00:18 | |
| *** changbl has quit IRC | 00:19 | |
| *** sanchitmalhotra1 has joined #openstack-swift | 00:23 | |
| *** sanchitmalhotra has quit IRC | 00:26 | |
| *** changbl has joined #openstack-swift | 00:29 | |
| *** sanchitmalhotra has joined #openstack-swift | 00:37 | |
| *** sanchitmalhotra1 has quit IRC | 00:39 | |
| *** sanchitmalhotra1 has joined #openstack-swift | 00:43 | |
| *** sanchitmalhotra has quit IRC | 00:45 | |
| *** Guest27579 has quit IRC | 00:50 | |
| *** mac_ified has quit IRC | 00:57 | |
| *** sanchitmalhotra has joined #openstack-swift | 01:20 | |
| *** sanchitmalhotra1 has quit IRC | 01:22 | |
| *** asettle is now known as asettle-afk | 01:34 | |
| *** sanchitmalhotra1 has joined #openstack-swift | 01:38 | |
| *** sanchitmalhotra has quit IRC | 01:40 | |
| *** haomaiwang has joined #openstack-swift | 01:41 | |
| *** jerrygb has joined #openstack-swift | 01:44 | |
| *** haomaiwang has quit IRC | 02:01 | |
| *** haomaiwang has joined #openstack-swift | 02:01 | |
| *** jerrygb has quit IRC | 02:05 | |
| *** jerrygb has joined #openstack-swift | 02:05 | |
| *** jerrygb has quit IRC | 02:10 | |
| *** sanchitmalhotra has joined #openstack-swift | 02:15 | |
| *** sanchitmalhotra1 has quit IRC | 02:17 | |
| *** asettle-afk is now known as asettle | 02:22 | |
| *** sanchitmalhotra1 has joined #openstack-swift | 02:38 | |
| *** sanchitmalhotra has quit IRC | 02:40 | |
| *** haomaiwang has quit IRC | 03:01 | |
| *** haomaiwang has joined #openstack-swift | 03:01 | |
| *** jerrygb has joined #openstack-swift | 03:06 | |
| *** venkat_p has joined #openstack-swift | 03:06 | |
| *** proteusguy_ has quit IRC | 03:08 | |
| *** jerrygb has quit IRC | 03:11 | |
| *** sanchitmalhotra has joined #openstack-swift | 03:22 | |
| *** sanchitmalhotra1 has quit IRC | 03:24 | |
| *** proteusguy_ has joined #openstack-swift | 03:25 | |
| *** sanchitmalhotra1 has joined #openstack-swift | 03:33 | |
| *** sanchitmalhotra has quit IRC | 03:35 | |
| *** changbl has quit IRC | 03:38 | |
| *** sanchitmalhotra has joined #openstack-swift | 03:41 | |
| *** sanchitmalhotra1 has quit IRC | 03:43 | |
| *** wuhg has joined #openstack-swift | 03:55 | |
| *** haomaiwang has quit IRC | 04:01 | |
| *** haomaiwang has joined #openstack-swift | 04:01 | |
| *** Zyric1 has quit IRC | 04:03 | |
| *** tsg_ has quit IRC | 04:06 | |
| *** diazjf has joined #openstack-swift | 04:07 | |
| *** changbl has joined #openstack-swift | 04:07 | |
| *** sanchitmalhotra has quit IRC | 04:10 | |
| *** ppai has joined #openstack-swift | 04:22 | |
| openstackgerrit | Matthew Oliver proposed openstack/swift: Port swift.common.utils.Timestamp to Python 3 https://review.openstack.org/237015 | 04:32 |
|---|---|---|
| *** venkat_p has quit IRC | 04:43 | |
| *** trifon has joined #openstack-swift | 04:49 | |
| *** haomaiwang has quit IRC | 05:01 | |
| *** trifon has quit IRC | 05:01 | |
| *** haomaiwa_ has joined #openstack-swift | 05:01 | |
| openstackgerrit | Merged openstack/swift: Ignore Content-Type from client on multipart-manifest=delete https://review.openstack.org/248186 | 05:06 |
| *** klrmn1 has quit IRC | 05:22 | |
| *** diazjf has quit IRC | 05:23 | |
| *** rcernin has joined #openstack-swift | 05:36 | |
| *** mwheckmann has joined #openstack-swift | 05:36 | |
| *** mwheckmann has quit IRC | 05:40 | |
| *** rcernin has quit IRC | 05:42 | |
| *** dimasot has joined #openstack-swift | 05:43 | |
| *** trifon has joined #openstack-swift | 05:44 | |
| *** diazjf has joined #openstack-swift | 05:48 | |
| *** haomaiwa_ has quit IRC | 06:01 | |
| *** 17WAA30H9 has joined #openstack-swift | 06:01 | |
| *** SkyRocknRoll has joined #openstack-swift | 06:05 | |
| *** silor has joined #openstack-swift | 06:06 | |
| *** kei_yama has quit IRC | 06:10 | |
| *** venkat_p has joined #openstack-swift | 06:14 | |
| *** mahatic has joined #openstack-swift | 06:38 | |
| *** silor1 has joined #openstack-swift | 06:46 | |
| *** eranrom has joined #openstack-swift | 06:47 | |
| *** silor has quit IRC | 06:48 | |
| *** silor1 is now known as silor | 06:48 | |
| *** mac_ified has joined #openstack-swift | 06:49 | |
| *** 17WAA30H9 has quit IRC | 07:01 | |
| *** haomaiwang has joined #openstack-swift | 07:01 | |
| *** pchng_ has joined #openstack-swift | 07:07 | |
| *** rcernin has joined #openstack-swift | 07:07 | |
| *** pchng has quit IRC | 07:09 | |
| *** openstack has joined #openstack-swift | 07:21 | |
| *** mahatic has joined #openstack-swift | 07:30 | |
| *** jerrygb has joined #openstack-swift | 07:51 | |
| *** mac_ified has quit IRC | 07:52 | |
| *** mac_ifie_ has joined #openstack-swift | 07:52 | |
| *** jerrygb has quit IRC | 07:57 | |
| *** hseipp has joined #openstack-swift | 08:00 | |
| *** haomaiwang has quit IRC | 08:01 | |
| *** haomaiwang has joined #openstack-swift | 08:01 | |
| *** hseipp has quit IRC | 08:11 | |
| *** diazjf has quit IRC | 08:11 | |
| *** arnox has joined #openstack-swift | 08:11 | |
| *** dimasot has quit IRC | 08:19 | |
| *** venkat_p has quit IRC | 08:31 | |
| *** haomaiwang has quit IRC | 08:32 | |
| openstackgerrit | Eran Rom proposed openstack/swift: Add thread level concurrency to container sync https://review.openstack.org/225338 | 08:33 |
| *** haomaiwang has joined #openstack-swift | 08:36 | |
| *** ppai_ has joined #openstack-swift | 08:40 | |
| *** ppai has quit IRC | 08:44 | |
| *** venkat_p has joined #openstack-swift | 08:49 | |
| *** jordanP has joined #openstack-swift | 08:52 | |
| *** haomaiwang has quit IRC | 09:01 | |
| *** haomaiwang has joined #openstack-swift | 09:01 | |
| *** mahatic has quit IRC | 09:08 | |
| *** SkyRocknRoll has quit IRC | 09:15 | |
| *** joeljwright has joined #openstack-swift | 09:25 | |
| *** ChanServ sets mode: +v joeljwright | 09:25 | |
| *** jistr has joined #openstack-swift | 09:27 | |
| *** aix has joined #openstack-swift | 09:43 | |
| *** SkyRocknRoll has joined #openstack-swift | 09:44 | |
| *** dimasot has joined #openstack-swift | 09:46 | |
| *** aix has quit IRC | 09:52 | |
| *** mahatic has joined #openstack-swift | 09:57 | |
| *** ekarlso- has joined #openstack-swift | 09:59 | |
| *** ekarlso has quit IRC | 10:00 | |
| *** ekarlso- is now known as ekarlso | 10:00 | |
| *** haomaiwang has quit IRC | 10:01 | |
| *** haomaiwang has joined #openstack-swift | 10:01 | |
| *** jerrygb has joined #openstack-swift | 10:08 | |
| *** jerrygb has quit IRC | 10:13 | |
| mahatic | acoles_: around? | 10:14 |
| *** daemontool has joined #openstack-swift | 10:19 | |
| *** aix has joined #openstack-swift | 10:24 | |
| *** Guest99452 has quit IRC | 10:33 | |
| *** daemontool has quit IRC | 10:34 | |
| *** daemontool has joined #openstack-swift | 10:35 | |
| openstackgerrit | Merged openstack/swift: Transition to 'keymaster' https://review.openstack.org/244904 | 10:53 |
| *** Fuz1on has joined #openstack-swift | 10:56 | |
| Fuz1on | Hello all | 10:56 |
| Fuz1on | is anybody there ? | 10:56 |
| *** hseipp has joined #openstack-swift | 10:56 | |
| Fuz1on | i got a question about log_custom_handlers, i dont really know where to put my logger ? | 10:57 |
| Fuz1on | i guess it's in /usr/lib/pythonX.X/dist-packages/swift | 10:57 |
| Fuz1on | but i'm not really sure | 10:57 |
| *** haomaiwang has quit IRC | 11:01 | |
| *** haomaiwang has joined #openstack-swift | 11:01 | |
| *** NM has joined #openstack-swift | 11:04 | |
| Fuz1on | i'll give it a try | 11:05 |
| Fuz1on | l will mimic the sentry logger | 11:05 |
| *** yatin has joined #openstack-swift | 11:05 | |
| *** haomaiwang has quit IRC | 11:07 | |
| *** jmccarthy has quit IRC | 11:24 | |
| *** jmccarthy has joined #openstack-swift | 11:25 | |
| openstackgerrit | Alistair Coles proposed openstack/swift: Decrypting Container Listing https://review.openstack.org/214438 | 11:34 |
| *** acoles_ is now known as acoles | 11:34 | |
| acoles | mahatic: hi | 11:35 |
| mahatic | acoles: hello! are you back? jusitce done? :) | 11:36 |
| acoles | mahatic: yes back to normal | 11:37 |
| mahatic | acoles: cool, do you have a few mins to chat about content-type encryption? | 11:38 |
| acoles | sure | 11:39 |
| acoles | mahatic: sure | 11:39 |
| mahatic | acoles: great. 1) crypto-meta is getting stripped off in response here: https://github.com/openstack/swift/blob/master/swift/common/swob.py#L403 | 11:39 |
| mahatic | acoles: I was wondering, if it is okay to put a condition there to check for "; meta=" in there and pass along if that is true | 11:40 |
| mahatic | acoles: for some context: I was working on patch 248377 and came across that issue | 11:42 |
| patchbot | mahatic: https://review.openstack.org/#/c/248377/ - Decrypt multi-part mime doc | 11:42 |
| *** yatin has quit IRC | 11:43 | |
| acoles | mahatic: :/ | 11:43 |
| acoles | mahatic: good catch | 11:44 |
| mahatic | acoles: another worry is: 2) for multipart-mime, when it hits this: https://github.com/openstack/swift/blob/master/swift/common/swob.py#L1099, it is still looking for encrypted content-type length (which will be diff after decryption) | 11:45 |
| mahatic | acoles: for more on content-type length: here is where it starts - https://github.com/openstack/swift/blob/master/swift/common/swob.py#L1181 | 11:46 |
| acoles | mahatic: back to 1) - I am not sure that the right solution is to test for 'meta' in _resp_content_type_property(). Instead, I would suggest that _resp_content_type_property() should explicitly remove only a param starting charset= | 11:50 |
| acoles | mahatic: but I'd like to understand where that method is being called that triggers the problem (I am sure it could be a problem!) | 11:50 |
| mahatic | acoles: which one? _resp_content_type_property()? | 11:51 |
| acoles | mahatic: yes. but i think i see it, line 1187?? | 11:52 |
| *** JelleB is now known as a1|away | 11:52 | |
| mahatic | acoles: https://github.com/openstack/swift/blob/master/swift/common/swob.py#L1116 | 11:52 |
| acoles | mahatic: yes. i meant where is the content_type property read - there's places that manipulate the resp content-type header and there's places that get/set the content-type property | 11:54 |
| mahatic | acoles: ok, in that case, yeah that would be 1187 | 11:55 |
| acoles | so is the problem you see that in the multipart boundary the content-type meta has been stripped off? because of line 1187 | 11:55 |
| acoles | yep! | 11:55 |
| acoles | mahatic: https://github.com/openstack/swift/blob/master/swift/common/swob.py#L412 also looks like it could cause problems | 11:58 |
| acoles | mahatic: those methods seem to assume that the only param is charset | 11:58 |
| acoles | mahatic: i wonder how swift_bytes avoids being stripped by same method? maybe its just luck that the content_type property isn't used? | 11:59 |
| mahatic | acoles: i think there is another method (or something that overwrites for swift_bytes. I *think*). Let me take a quick look | 12:00 |
| mahatic | I remember coming across that | 12:00 |
| acoles | there is | 12:01 |
| mahatic | acoles: ok, and here in response, weŕe only reading _resp_charset_property llooking specifically for charset I believe? I didnt think this would cause a problem | 12:04 |
| mahatic | and looking* | 12:04 |
| *** SkyRocknRoll has quit IRC | 12:09 | |
| mahatic | override_bytes_from_content_type - i was looking for this one earlier | 12:11 |
| *** jmccarthy has quit IRC | 12:11 | |
| *** [1]dimasot has joined #openstack-swift | 12:13 | |
| acoles | mahatic: I even wonder why the charset param is stripped off the content-types in the multipart doc? maybe that is not intentional? the example here shows charset included http://www.w3.org/Protocols/rfc1341/7_2_Multipart.html | 12:15 |
| *** dimasot has quit IRC | 12:17 | |
| *** [1]dimasot is now known as dimasot | 12:17 | |
| mahatic | acoles: looking | 12:17 |
| *** jmccarthy has joined #openstack-swift | 12:19 | |
| acoles | mahatic: i.e. i am thinking that here https://github.com/openstack/swift/blob/master/swift/common/swob.py#L1187 it should actually read the content-type from the response headers, NOT via the content_type property, so that the charset is not lost. Otherwise AFAICT the response loses any record of charset. | 12:20 |
| mahatic | acoles: maybe the intention is that it sets here? https://github.com/openstack/swift/blob/master/swift/common/swob.py#L1158 | 12:22 |
| *** SkyRocknRoll has joined #openstack-swift | 12:25 | |
| acoles | mahatic: yes, but the problem i see is the use of the Response property attribute to retrieve content-type when creating the multipart doc. swob is swift's version of webob and swob.Response.content_type appears to do the right thing according to http://docs.webob.org/en/1.1/modules/webob.html#response. But I'm not sure the charset should be dropped from the multipart doc part. | 12:26 |
| acoles | mahatic: so my assertion is that line 1187 in swob should read content_type = self.headers.get('content-type') | 12:29 |
| acoles | mahatic: which would mean that both charset and crypto meta would appear in the multipart doc part | 12:30 |
| acoles | mahatic: i'll try asking torgomatic later. i'm sure he'll know The Right Thing :) | 12:31 |
| mahatic | acoles: :) ok. but what is L1131 doing? (!) | 12:32 |
| mahatic | acoles: is it setting the content-type header? but i thought we are retrieving and stripping things off and everything | 12:33 |
| acoles | mahatic: line 1131 is setting a default value. then see line 1150, updates with supplied values | 12:34 |
| *** marcusvrn_ has joined #openstack-swift | 12:35 | |
| mahatic | acoles: using self.headers.get('content-type') in multipart makes sense. but then we are also calculating length there https://github.com/openstack/swift/blob/master/swift/common/swob.py#L1196 which will be length of encrypted value | 12:36 |
| acoles | mahatic: line 1196 is calculating the length of whatever is going to be written as the multipart boundary. So yes, that should include the length of the encrypted content-type and its meta param | 12:38 |
| mahatic | acoles: and somehow line 1103 is holding onto that value - i.e. in the decrypter response, it doesnt move forward if the length is not equivalent to encrypted value(or whatever was before) | 12:39 |
| mahatic | :/ | 12:39 |
| mahatic | acoles: or maybe that is my mistake, and it is not supposed to be that way? as in on decrypter response, this whole response thing will go through the process of putting together those values again? | 12:40 |
| acoles | mahatic: yep, my expectation is that the decrypter will need to rebuild the multipart response body with new multipart boundaries each having decrypted content-types | 12:41 |
| mahatic | acoles: it does that currently (as in the output of my code), but except the content-type, expecting the encrypted length. Will check that out | 12:41 |
| *** lpabon has joined #openstack-swift | 12:42 | |
| acoles | mahatic: there are some helper methods in utils that may be useful for reading the multipart body e.g. swift/common/utils.py:3477 | 12:45 |
| *** cdelatte has joined #openstack-swift | 12:45 | |
| mahatic | acoles: running patch 248377 against patch 167828 passes, except for content-type assertion | 12:46 |
| patchbot | mahatic: https://review.openstack.org/#/c/248377/ - Decrypt multi-part mime doc | 12:46 |
| patchbot | mahatic: https://review.openstack.org/#/c/167828/ - Add functional test for multi-range GET requests. | 12:46 |
| *** aix has quit IRC | 12:47 | |
| *** aix has joined #openstack-swift | 12:48 | |
| acoles | mahatic: but if the crypto meta was being lost from the content-type then it wouldn't get decrypted so you'd expect that to assertion to fail, no? | 12:48 |
| mahatic | acoles: and i just ran the code putting some print in swob, inside those content_type and length stuff - interestingly it only goes through once, that showing up with encrypted values | 12:48 |
| mahatic | acoles: oh yeah, that was expected. just saying that i used those helper methods you were pointing out and i do have a multipart-doc output, with data decrypted (which the test asserts) | 12:49 |
| acoles | mahatic: cool. so maybe try chaning swob line 1187 to content_type = self.headers.get('content-type') ?? | 12:50 |
| acoles | *changing | 12:50 |
| mahatic | acoles: i also did that (not exactly self.headers, but as a temp workaround by filtering ; meta) and *then* I encountered the length issue that im talking about now :) | 12:51 |
| acoles | mahatic: oic | 12:53 |
| mahatic | acoles: im not sure if the entire response is constructed again i.e. all those values are processed again for decrypter, I was thinking, decrypter takes the response and modifies and then processes the same way it got (same values for response stuff) | 12:54 |
| mahatic | acoles: sorry, by "all those values are processed again for decrypter" - i mean all those values that "Response" is holding... Well it still doesnt sound to make a lot of sense, but i only Response being called once, in this whole process :/ | 12:56 |
| mahatic | only see* | 12:56 |
| *** jerrygb has joined #openstack-swift | 12:59 | |
| *** jerrygb has quit IRC | 12:59 | |
| *** jerrygb has joined #openstack-swift | 12:59 | |
| *** haomaiwa_ has joined #openstack-swift | 12:59 | |
| *** jerrygb_ has joined #openstack-swift | 13:00 | |
| *** haomaiwa_ has quit IRC | 13:01 | |
| *** haomaiwang has joined #openstack-swift | 13:01 | |
| *** jerrygb has quit IRC | 13:03 | |
| acoles | mahatic: are you going to update your patch with any of the changes discussed here? or is it up to date? | 13:03 |
| acoles | mahatic: away for lunch | 13:04 |
| mahatic | acoles: I need to replace the self.get.headers and push it, will do so (And that will leave me with the length issue, trying to figure it out) | 13:05 |
| mahatic | acoles: sure, have a good lunch | 13:06 |
| mahatic | replace with* | 13:06 |
| *** SkyRocknRoll has quit IRC | 13:08 | |
| *** ppai_ has quit IRC | 13:11 | |
| mahatic | acoles: i think i figured the length issue - i dont think im updating content-length value | 13:21 |
| *** SkyRocknRoll has joined #openstack-swift | 13:21 | |
| mahatic | acoles: thanks for looking into this! | 13:21 |
| *** haomaiwang has quit IRC | 13:22 | |
| *** haomaiwa_ has joined #openstack-swift | 13:24 | |
| *** daemontool has quit IRC | 13:25 | |
| *** venkat_p has quit IRC | 13:25 | |
| acoles | mahatic: ok, sounds like progress | 13:32 |
| *** SkyRocknRoll has quit IRC | 13:32 | |
| mahatic | acoles: sounds so | 13:33 |
| *** daemontool has joined #openstack-swift | 13:38 | |
| eranrom | acoles: I am reviewing the rbac patch (202411), and wondering where is the policy.json from which it is derived (it is derived from some policy.json right?) | 13:44 |
| *** jistr has quit IRC | 13:46 | |
| acoles | eranrom: hi! i'm not sure if ho used a policy.json file to derive the scenarios. i remember he had a description of his methodology in a pastebin. | 13:58 |
| acoles | eranrom: http://paste.openstack.org/show/412312/ and links in that page | 13:58 |
| acoles | eranrom: i'm not sure if that is ho's 'top level' methodology description ^^ | 13:59 |
| *** haomaiwa_ has quit IRC | 14:01 | |
| *** haomaiwang has joined #openstack-swift | 14:01 | |
| eranrom | acoles: Thanks. will start from there (^^) | 14:04 |
| acoles | eranrom: maybe you would catch ho in channel early in your daytime | 14:04 |
| *** tongli has joined #openstack-swift | 14:05 | |
| eranrom | acoles: yep will try that tomorrow | 14:05 |
| *** daemontool has quit IRC | 14:12 | |
| *** daemontool has joined #openstack-swift | 14:13 | |
| *** blmartin has joined #openstack-swift | 14:22 | |
| *** pchng_ is now known as pchng | 14:23 | |
| *** hseipp has quit IRC | 14:31 | |
| *** hseipp has joined #openstack-swift | 14:33 | |
| *** NM has quit IRC | 14:42 | |
| *** j_king_ is now known as j_king | 14:42 | |
| *** janonymous__ has joined #openstack-swift | 14:43 | |
| *** mzhou has joined #openstack-swift | 14:47 | |
| *** zacksh_ is now known as zacksh | 14:47 | |
| *** aix has quit IRC | 14:54 | |
| *** petertr7 is now known as petertr7_away | 14:57 | |
| *** haomaiwang has quit IRC | 15:01 | |
| *** Asselin has joined #openstack-swift | 15:01 | |
| *** haomaiwa_ has joined #openstack-swift | 15:01 | |
| *** Asselin has left #openstack-swift | 15:01 | |
| *** aix has joined #openstack-swift | 15:01 | |
| *** openstackgerrit has quit IRC | 15:02 | |
| *** openstackgerrit has joined #openstack-swift | 15:02 | |
| *** petertr7_away is now known as petertr7 | 15:02 | |
| *** mwheckmann has joined #openstack-swift | 15:03 | |
| *** admin6 has joined #openstack-swift | 15:05 | |
| *** breitz has joined #openstack-swift | 15:08 | |
| *** foexle has joined #openstack-swift | 15:12 | |
| *** pchng_ has joined #openstack-swift | 15:15 | |
| *** pchng has quit IRC | 15:16 | |
| *** wbhuber_ has joined #openstack-swift | 15:17 | |
| *** wbhuber_ has quit IRC | 15:17 | |
| *** pdardeau has joined #openstack-swift | 15:17 | |
| *** pdardeau has quit IRC | 15:22 | |
| *** pdardeau has joined #openstack-swift | 15:23 | |
| *** mwheckmann has quit IRC | 15:27 | |
| *** mzhou has quit IRC | 15:28 | |
| *** willsama has joined #openstack-swift | 15:31 | |
| *** willsama has quit IRC | 15:31 | |
| *** daddyjoseph97 has joined #openstack-swift | 15:35 | |
| *** mwheckmann has joined #openstack-swift | 15:36 | |
| *** NM has joined #openstack-swift | 15:39 | |
| *** jlhinson has joined #openstack-swift | 15:42 | |
| *** daemontool_ has joined #openstack-swift | 15:48 | |
| *** jlhinson has quit IRC | 15:48 | |
| *** daemontool has quit IRC | 15:52 | |
| *** NM has quit IRC | 15:57 | |
| *** haomaiwa_ has quit IRC | 16:01 | |
| *** foexle has quit IRC | 16:01 | |
| *** haomaiwa_ has joined #openstack-swift | 16:01 | |
| *** mwheckmann has quit IRC | 16:01 | |
| *** blmartin has quit IRC | 16:03 | |
| *** petertr7 is now known as petertr7_away | 16:06 | |
| *** rcernin has quit IRC | 16:08 | |
| *** Pablo|off| has joined #openstack-swift | 16:14 | |
| *** Pablo|off| is now known as pcaruana | 16:15 | |
| *** eranrom has quit IRC | 16:16 | |
| *** hseipp has quit IRC | 16:17 | |
| *** hseipp has joined #openstack-swift | 16:17 | |
| *** pchng_ is now known as pchng | 16:20 | |
| *** petertr7_away is now known as petertr7 | 16:21 | |
| *** hseipp has quit IRC | 16:22 | |
| *** SkyRocknRoll has joined #openstack-swift | 16:24 | |
| *** aix has quit IRC | 16:24 | |
| *** klrmn1 has joined #openstack-swift | 16:26 | |
| *** blmartin has joined #openstack-swift | 16:26 | |
| *** diazjf has joined #openstack-swift | 16:27 | |
| *** diazjf1 has joined #openstack-swift | 16:28 | |
| *** klrmn1 has quit IRC | 16:30 | |
| *** trifon has quit IRC | 16:30 | |
| *** diazjf has quit IRC | 16:31 | |
| *** jistr has joined #openstack-swift | 16:33 | |
| notmyname | good morning | 16:41 |
| joeljwright | morning | 16:41 |
| pdardeau | morning | 16:41 |
| notmyname | it's cool and foggy this am in san francisco. makes it hard to get started in the morning :-) | 16:42 |
| notmyname | reminder that there is NOT a meeting this week | 16:42 |
| peluse | morning | 16:43 |
| *** petertr7 is now known as petertr7_away | 16:44 | |
| *** mwheckmann has joined #openstack-swift | 16:46 | |
| mahatic | good morning | 16:47 |
| mahatic | notmyname: we all can sympathise with that :) | 16:48 |
| *** joeljwright1 has joined #openstack-swift | 16:48 | |
| *** petertr7_away is now known as petertr7 | 16:49 | |
| *** mwheckmann has quit IRC | 16:51 | |
| *** joeljwright has quit IRC | 16:52 | |
| peluse | I dunno, its pretty nice here in AZ :) | 16:53 |
| *** mwheckmann has joined #openstack-swift | 16:55 | |
| notmyname | looks like today* is the day that the gate is killing py26 support | 16:55 |
| notmyname | * maybe not today | 16:56 |
| notmyname | or rather, today oslo is dropping py26 and doing a major version bump and it's likely the gate stuff will be remove within a week | 16:56 |
| notmyname | so, whatever we have in python-swiftclient now is pretty close to what we will have in a last release before we bump the version and do a py26-free release | 16:57 |
| joeljwright1 | sounds reasonable | 16:57 |
| notmyname | joeljwright1: timburke: is there any open patch that we really shouldn't do a release without landing it first? | 16:59 |
| *** haomaiwa_ has quit IRC | 17:01 | |
| *** 21WAAEGOI has joined #openstack-swift | 17:01 | |
| joeljwright1 | notmyname: not that I can think of | 17:01 |
| joeljwright1 | I've just had a look at the list | 17:01 |
| notmyname | I'm looking now too | 17:01 |
| notmyname | thanks | 17:01 |
| notmyname | joeljwright1: are you still working on patch 223319? | 17:03 |
| patchbot | notmyname: https://review.openstack.org/#/c/223319/ - New API documentation for python-swiftclient | 17:03 |
| joeljwright1 | notmyname: yes | 17:03 |
| joeljwright1 | it still has a lot of work remaining | 17:03 |
| joeljwright1 | and I was hoping to get the suggestions from tokyo | 17:04 |
| notmyname | ok. is there something that can be landed sooner? smaller chunks? | 17:04 |
| notmyname | ah yes. that. I'm still working on it too ;-) | 17:04 |
| joeljwright1 | it's possible it can be split up | 17:04 |
| joeljwright1 | but tbh I was using that patch as a way of getting suggestions | 17:05 |
| joeljwright1 | I'm unlikely to get any more work done on it before the end of December because of my trip | 17:05 |
| notmyname | ok | 17:05 |
| joeljwright1 | I can probably just complete the sections I have planned as phase 1 | 17:06 |
| joeljwright1 | (that is, complete the empty sections that are there now) | 17:06 |
| notmyname | joeljwright1: that would be great. also, if you don't have somethign smaller, then a note in the comments about someone else picking it up would be good | 17:06 |
| *** daemontool_ has quit IRC | 17:06 | |
| notmyname | joeljwright1: could you look at patch 230100? | 17:07 |
| patchbot | notmyname: https://review.openstack.org/#/c/230100/ - Add py35 to default tox environments | 17:07 |
| joeljwright1 | 1-liners that we obviously need | 17:08 |
| joeljwright1 | I like those | 17:08 |
| *** daemontool has joined #openstack-swift | 17:08 | |
| *** ir2ivps8 has joined #openstack-swift | 17:13 | |
| joeljwright1 | notmyname: the biggest question I have with the docs is what to do about the swiftclient.client docs | 17:16 |
| joeljwright1 | I still have a few code examples to create for swiftclient.service | 17:16 |
| joeljwright1 | but there is still a lack of guidance on how to choose client vs service vs cli | 17:17 |
| joeljwright1 | and nothing beyond the docstrings for client | 17:17 |
| notmyname | so what's the question? | 17:18 |
| *** sgundur has joined #openstack-swift | 17:19 | |
| joeljwright1 | what do we want to do about producing extra docs for swiftclient.client? Expand the docstrings? Produce another file with examples? | 17:19 |
| *** 21WAAEGOI has quit IRC | 17:20 | |
| notmyname | I definitely think we should, of course :-) | 17:20 |
| joeljwright1 | and how much do we want to write about picking the correct api to use? | 17:20 |
| notmyname | I'm not sure of the best way. docstrings and separate docs are both good. really, it probably doesn't matter much, compared to what we have now (or rather what we don't have) | 17:20 |
| *** rcernin has joined #openstack-swift | 17:21 | |
| notmyname | picking the correct API is the first place everyone should start | 17:21 |
| notmyname | is it's got to be good | 17:21 |
| joeljwright1 | indeed, and the second place most people fall down is auth | 17:21 |
| notmyname | yup | 17:22 |
| joeljwright1 | and the big issue with auth it that I'm fairly sure there is no one person who understands every path! | 17:22 |
| joeljwright1 | I'm as guilty as most of knowing how to auth with my systems, but not really what all the options are | 17:23 |
| *** admin6 has quit IRC | 17:23 | |
| joeljwright1 | and trying to get people involved in doc writing is… tricky… | 17:23 |
| joeljwright1 | :D | 17:24 |
| *** blmartin has quit IRC | 17:24 | |
| joeljwright1 | ok, I shall commit to trying to get my patch into a decent state before the new year | 17:25 |
| joeljwright1 | and I'll try to enlist some help with the auth and choosing parts | 17:26 |
| *** jerrygb_ has quit IRC | 17:26 | |
| *** janonymous__ has quit IRC | 17:26 | |
| joeljwright1 | (but I'm away on Friday until Dec 12th, so I shall be starting after that) | 17:27 |
| notmyname | thanks | 17:28 |
| *** petertr7 is now known as petertr7_away | 17:29 | |
| *** NM has joined #openstack-swift | 17:33 | |
| *** jerrygb has joined #openstack-swift | 17:33 | |
| *** travisn has joined #openstack-swift | 17:35 | |
| notmyname | https://pypi.python.org/pypi/django-storage-swift/1.2.5 was just updated. I love seeing stuff like that out there :-) | 17:35 |
| *** jerrygb has quit IRC | 17:36 | |
| *** CaioBrentano has joined #openstack-swift | 17:39 | |
| *** jerrygb has joined #openstack-swift | 17:40 | |
| *** petertr7_away is now known as petertr7 | 17:41 | |
| *** aix has joined #openstack-swift | 17:45 | |
| *** jerrygb has quit IRC | 17:45 | |
| openstackgerrit | Alistair Coles proposed openstack/swift: Fix listing of SLO manifest bytes after POST or COPY https://review.openstack.org/248219 | 17:46 |
| *** sc68cal has quit IRC | 17:46 | |
| acoles | jrichli: ^^ it got more complicated since you mentioned POSTs in your review ;) | 17:47 |
| *** sc68cal has joined #openstack-swift | 17:47 | |
| *** blmartin has joined #openstack-swift | 17:49 | |
| acoles | torgomatic: are you here? | 17:52 |
| notmyname | acoles: I didn't see him at his desk just now. might be commuting currently | 17:53 |
| *** zul has quit IRC | 17:53 | |
| acoles | notmyname: thanks. so turkey-fest hasn't started yet? | 17:53 |
| notmyname | the official day is thursday. but different families do different things | 17:54 |
| notmyname | starts for me very early on wednesday when we get on a plane | 17:54 |
| acoles | notmyname: yeah i imagine a lot of folks travel this week | 17:55 |
| *** eranrom has joined #openstack-swift | 17:56 | |
| *** NM has quit IRC | 17:57 | |
| *** zul has joined #openstack-swift | 17:58 | |
| *** CaioBren_ has joined #openstack-swift | 17:59 | |
| openstackgerrit | Alistair Coles proposed openstack/swift: Update .mailmap entry https://review.openstack.org/248202 | 18:02 |
| *** CaioBrentano has quit IRC | 18:03 | |
| notmyname | acoles: thanks. checking it now | 18:03 |
| *** joeljwright1 has quit IRC | 18:04 | |
| notmyname | acoles: looks good for what's there. but I see 4 more people in `git shortlog` that have hp.com addresses. Dragos, Dan, Jola, and Thomas | 18:05 |
| notmyname | acoles: I'm fine with approving what you have now, but do you know about those 4? | 18:05 |
| acoles | notmyname: i was going to add a comment but you beat me - three i now for sure are no longer @hpe.com i.e. Dan, Jola, and Thomas | 18:06 |
| notmyname | ok | 18:06 |
| notmyname | then let's leave it as-is for now | 18:07 |
| notmyname | it's not like it's hard to change the .mailmap later ;-) | 18:07 |
| notmyname | thanks for doing the research | 18:07 |
| notmyname | acoles: FWIW, this makes the AUTHORS file somewhat out of date, but I'll square that up with one commit myself when I add all the new contributors | 18:07 |
| peluse | notmyname: is there a trick to getting patch 214206 to pass all func tests? I've upgraded all my requirements and test-requirements and get a bunch of failures running func tests on this patch | 18:08 |
| patchbot | peluse: https://review.openstack.org/#/c/214206/ - Modify functional tests to use testr | 18:08 |
| acoles | notmyname: also i assume Dragos is not at hpe.com since the mail bounced, i just don't know them personally | 18:08 |
| notmyname | peluse: locally? passed for me locally, but maybe what you're seeing is similar to what the gate sees | 18:08 |
| acoles | notmyname: ah. you want me to update AUTHORS at same time (same review)? | 18:09 |
| peluse | yeah, locally. FAILED (failures=33, skipped=12) | 18:09 |
| notmyname | acoles: no. don't worry about it. I'll do it later when I update it for a release | 18:09 |
| acoles | notmyname: ok | 18:09 |
| notmyname | acoles: I've got a script which tells me what's different between shortlog and the AUTHORS frile | 18:09 |
| *** arnox has quit IRC | 18:10 | |
| notmyname | peluse: trying it again locally | 18:14 |
| peluse | notmyname: cool, I'll try again too. maybe it will work the 3nd or 3rd time :) | 18:16 |
| peluse | notmyname: I know my pip/pbr/setuptols are all latest/greatest and I went ahead and did --upgrade on both requirements and the updated test-requirements as well. Is that not a good thing to do? | 18:17 |
| *** jordanP has quit IRC | 18:17 | |
| *** aix has quit IRC | 18:18 | |
| notmyname | they had to be updated for other reasons lately. shouldn't be tied particularly to this specific patch though | 18:19 |
| *** klrmn1 has joined #openstack-swift | 18:20 | |
| peluse | notmyname: thats what I thought but I had some obvious failures with imports/etc with whatever state all my dependences were in so --upgraded everything. Til I did thtat they wouldnt even execute | 18:20 |
| timburke | notmyname: on python-swiftclient, i'd kinda like to see https://review.openstack.org/#/c/190887/ land (since it give you a means to post data to an account and inspect the response), but there's enough going on there that it probably wouldn't be a quick review | 18:21 |
| acoles | peluse: thanks for your review on patch 231121, i'm back from hanging duty and will respond to your comments 'real soon' | 18:21 |
| patchbot | acoles: https://review.openstack.org/#/c/231121/ - Make ECDiskFile report all fragments found on disk | 18:21 |
| peluse | acoles: awesome! hope you didn't send any innocents to the gallows :) | 18:22 |
| *** CaioBren_ has quit IRC | 18:23 | |
| acoles | peluse: hmmm, its kinda hard to know :) | 18:23 |
| peluse | acoles: just kidding, if they were accused I'm sure they were guilty! | 18:23 |
| *** CaioBrentano has joined #openstack-swift | 18:23 | |
| notmyname | *shudder* | 18:23 |
| * peluse wonders why he's never picked for jury duty... | 18:23 | |
| notmyname | Ran 380 tests in 305.450s | 18:24 |
| notmyname | OK (skipped=12) | 18:24 |
| notmyname | peluse: ^ | 18:24 |
| notmyname | peluse: what's failing on your machined? | 18:24 |
| peluse | notmyname: wow | 18:25 |
| acoles | peluse: heh. | 18:25 |
| peluse | notmyname: ton of stuff. running now, when its done I'll pick out a few and psate them in | 18:25 |
| peluse | gmmaha: how many of your func tests are failing? | 18:26 |
| *** jerrygb has joined #openstack-swift | 18:27 | |
| *** mahatic has quit IRC | 18:27 | |
| peluse | any core: pls take a mn to look at a quick code coverage patch in need of 2nd +2. patch 240613 | 18:28 |
| patchbot | peluse: https://review.openstack.org/#/c/240613/ - Unit tests for account/backend.py | 18:28 |
| notmyname | patch 245266 really should land before a release | 18:28 |
| patchbot | notmyname: https://review.openstack.org/#/c/245266/ - remove pbr from requirements.txt. It's not a run-t... | 18:28 |
| notmyname | peluse: looking at that one | 18:29 |
| openstackgerrit | Alistair Coles proposed openstack/swift: Stop staticweb revealing container existence to unauth'd requests https://review.openstack.org/248867 | 18:29 |
| peluse | notmyname: gracias | 18:30 |
| *** NM has joined #openstack-swift | 18:35 | |
| *** petertr7 is now known as petertr7_away | 18:41 | |
| *** wuhg has quit IRC | 18:41 | |
| *** jerrygb has quit IRC | 18:42 | |
| clayg | heyoh! | 18:42 |
| peluse | hola! | 18:45 |
| *** NM has quit IRC | 18:46 | |
| *** jerrygb has joined #openstack-swift | 18:46 | |
| openstackgerrit | Merged openstack/python-swiftclient: Add py35 to default tox environments https://review.openstack.org/230100 | 18:46 |
| peluse | notmyname: here's my fnc test failures on the testr thing: https://gist.github.com/anonymous/e036f1bab957ac81fde8 | 18:48 |
| * peluse BRB | 18:49 | |
| *** wbhuber_ has joined #openstack-swift | 18:49 | |
| notmyname | in coverage reports, what's the difference between a "missing" line statement of 1234-5678 and 1234->5678? | 18:50 |
| notmyname | what's the difference between the - and -> | 18:50 |
| *** acoles is now known as acoles_ | 18:50 | |
| *** jerrygb has quit IRC | 18:51 | |
| notmyname | with patch 240613, I get 608->609 in the missing column. with master I get 607-609. the patch should be adding coverage for 608 and 609 | 18:52 |
| patchbot | notmyname: https://review.openstack.org/#/c/240613/ - Unit tests for account/backend.py | 18:52 |
| notmyname | peluse: ^ | 18:52 |
| *** daemontool has quit IRC | 18:54 | |
| *** trifon has joined #openstack-swift | 18:54 | |
| *** jerrygb has joined #openstack-swift | 19:03 | |
| *** wbhuber_ has quit IRC | 19:08 | |
| *** diazjf1 has quit IRC | 19:11 | |
| gmmaha | peluse: i got 75 failed | 19:11 |
| gmmaha | notmyname: for 608 and 609, i was never able to very that 609 is being raised | 19:14 |
| gmmaha | i was able to test for 608, but then 609 was always a puzzle.. had a similar one in the same file upto which i couldnt fix either | 19:14 |
| *** silor has quit IRC | 19:17 | |
| *** diazjf has joined #openstack-swift | 19:28 | |
| *** petertr7_away is now known as petertr7 | 19:45 | |
| gmmaha | notmyname: found where i was going wrong.. fixing the patch now.. | 20:00 |
| *** jordanP has joined #openstack-swift | 20:03 | |
| *** jordanP has quit IRC | 20:05 | |
| notmyname | gmmaha: ah, great | 20:11 |
| *** mwheckmann has quit IRC | 20:17 | |
| openstackgerrit | Merged openstack/swift: Added unit test cases for container/auditor.py https://review.openstack.org/240351 | 20:17 |
| openstackgerrit | Merged openstack/swift: Update .mailmap entry https://review.openstack.org/248202 | 20:19 |
| *** jistr has quit IRC | 20:23 | |
| *** CaioBren_ has joined #openstack-swift | 20:34 | |
| *** CaioBren_ has quit IRC | 20:36 | |
| *** CaioBrentano has quit IRC | 20:37 | |
| *** petertr7 is now known as petertr7_away | 20:43 | |
| *** petertr7_away is now known as petertr7 | 20:46 | |
| *** Zyric1 has joined #openstack-swift | 20:52 | |
| *** SkyRocknRoll has quit IRC | 20:53 | |
| *** jrichli has joined #openstack-swift | 20:54 | |
| openstackgerrit | Ganesh Maharaj Mahalingam proposed openstack/swift: Unit tests for account/backend.py https://review.openstack.org/240613 | 21:10 |
| *** jodah- has quit IRC | 21:11 | |
| *** anteaya has quit IRC | 21:11 | |
| gmmaha | notmyname: is it ok if we re-arrange the import section to be alphabetical order? | 21:13 |
| notmyname | acoles_: you* broke my dashboard | 21:13 |
| gmmaha | purely cosmetic and also will help us comply with a pep8 rule i think | 21:13 |
| notmyname | acoles_: *actually the change from hp.com to hpe.com broke the dashboard | 21:13 |
| notmyname | gmmaha: it's not a pep8/flake/hacking rule we check, so there won't be any consistency or enforcement to keep it in whatever order you put it in | 21:14 |
| *** travisn has quit IRC | 21:14 | |
| notmyname | gmmaha: I'd suggest leaving it at-is | 21:14 |
| gmmaha | notmyname: ok.. Will let it be.. thanks | 21:15 |
| gmmaha | was more thinking we can remove it from the exceptions list and start checking for that rule.. ;) | 21:15 |
| gmmaha | but then its just my OCD!! | 21:15 |
| *** jodah has joined #openstack-swift | 21:21 | |
| *** jerrygb has quit IRC | 21:22 | |
| *** jerrygb has joined #openstack-swift | 21:26 | |
| *** anteaya has joined #openstack-swift | 21:26 | |
| *** travisn has joined #openstack-swift | 21:28 | |
| *** jerrygb has quit IRC | 21:31 | |
| *** david-lyle has quit IRC | 21:31 | |
| *** daddyjoseph97 has quit IRC | 21:32 | |
| *** david-lyle has joined #openstack-swift | 21:34 | |
| *** jrichli has quit IRC | 21:36 | |
| Zyric1 | Hi everyone! I'm returning to a patch I started a few weeks ago https://review.openstack.org/#/c/241978/ and have created a new SAIO for it and inserted the changed files. I have tried running tox -e func, .unittests, .functtests, and .probetest and am having trouble re-creating the error jenkins found. Could anyone please advise me how to run the exact tests which failed the patch please? | 21:41 |
| notmyname | hello Zyric1 | 21:42 |
| Zyric1 | Hi notmyname :) | 21:42 |
| *** diazjf has quit IRC | 21:42 | |
| notmyname | Zyric1: do you know how to get the logs from jenkins to see what failed there? | 21:42 |
| mattoliverau | Morning | 21:43 |
| notmyname | mattoliverau: hi. I was just looking into geelong accommodation. I need to figure out some other logistics before I book a room/flight, though | 21:44 |
| notmyname | mattoliverau: seems like deakin univ is really small. is that the same location it was at last time? the maps makes it seem like it's just one square block | 21:44 |
| notmyname | Zyric1: if not, I'll tell you how. just don't want to skip ahead in getting a solution | 21:45 |
| Zyric1 | notmyname: To investigate I clicked the link of the failure and read the contents of console.html.gz which seems to suggest some kind of mapping issue with the parameters to config. I wanted to have a play around in the code but I can't recreate the problem locally. | 21:45 |
| notmyname | yeah | 21:45 |
| notmyname | so the error is at http://logs.openstack.org/78/241978/3/check/gate-swift-tox-func/e0994e6/console.html.gz#_2015-11-10_09_31_23_298 | 21:45 |
| mattoliverau | notmyname: We've found some hotel accommodation, let me get the name of one were using | 21:45 |
| notmyname | mattoliverau: oh, good. thanks | 21:45 |
| notmyname | Zyric1: and it's interesting that it failed in the gate and not in the qa cluster | 21:48 |
| notmyname | hmm... /me as an idea | 21:48 |
| gmmaha | ls -lrt | 21:53 |
| notmyname | Zyric1: I duplicated it | 21:53 |
| notmyname | just now | 21:53 |
| notmyname | Zyric1: ...a little more detective work now... | 21:54 |
| Zyric1 | notmyname: Ah, nice! Which commands/tests did you use to duplicate it? | 21:54 |
| *** david-lyle has quit IRC | 21:55 | |
| *** zaitcev has joined #openstack-swift | 21:56 | |
| *** ChanServ sets mode: +v zaitcev | 21:56 | |
| openstackgerrit | Tim Burke proposed openstack/swift: Reverse-listings follow-up https://review.openstack.org/248924 | 21:56 |
| mattoliverau | notmyname: just got word, we're staying at the Bayside Geelong. | 21:59 |
| notmyname | mattoliverau: thanks | 21:59 |
| mattoliverau | I think that was a good price point to location | 21:59 |
| *** sgundur has quit IRC | 21:59 | |
| *** david-lyle has joined #openstack-swift | 22:03 | |
| *** david-lyle has quit IRC | 22:04 | |
| *** sgundur has joined #openstack-swift | 22:05 | |
| *** petertr7 is now known as petertr7_away | 22:06 | |
| notmyname | Zyric1: ok, so it's using the "in-process" fuctional tests. which are a little weird, admittedly, but what's happening is that there's a swift.conf file being generated that has the line "aliases = ['Policy-0']" and that's abviously wrong | 22:09 |
| notmyname | I'm not sure yet how that's getting stuck in there | 22:09 |
| notmyname | Zyric1: duplicate it with `SWIFT_TEST_IN_PROCESS=true SWIFT_TEST_IN_PROCESS_CONF_DIR=/tmp tox -efunc` | 22:09 |
| *** jerrygb has joined #openstack-swift | 22:11 | |
| notmyname | Zyric1: fyi search for "in-process" on http://docs.openstack.org/developer/swift/development_guidelines.html | 22:12 |
| Zyric1 | notmyname: Many thanks, I'll look into that now. | 22:19 |
| notmyname | Zyric1: here's what i kno | 22:19 |
| notmyname | the storage policies are being loaded twice | 22:20 |
| *** rcernin has quit IRC | 22:20 | |
| *** gmmaha has quit IRC | 22:20 | |
| notmyname | the first time, it's working ok, but I think it's creating the aliases attribute and the second load is making that get written out into a config file | 22:20 |
| notmyname | and str(list) ends up being the value ['Policy-0'] | 22:20 |
| notmyname | which fails the parsing | 22:21 |
| *** david-lyle has joined #openstack-swift | 22:21 | |
| *** gmmaha has joined #openstack-swift | 22:21 | |
| notmyname | so I don't like the double loading, but i'm not sure if that's the problem or not. or at least if it is a problem it's not the only one | 22:21 |
| notmyname | part of the issue, i think, is that the aliases attribute is a list and that's getting serialized to a string | 22:22 |
| notmyname | maybe there needs to be a "human readable or config" attribute as a string and another as a list for parsing | 22:22 |
| notmyname | maybe the patch simply needs to be more careful about when the aliases value is used | 22:23 |
| clayg | notmyname: this is something that's broke or something we're trying to fix | 22:24 |
| notmyname | clayg: depends on what's broke ;-) | 22:25 |
| *** david-lyle has quit IRC | 22:25 | |
| notmyname | I'm not sure if the double loading thing is what's broken or if it's the erroneous addition of the value into the config file. or both. | 22:26 |
| Zyric1 | notmyname: Thanks for all the information, that's super useful. Is the double loading an intentional thing? If so to what end? I'll have a think and tinker with the aliases bit and see where that leads. | 22:28 |
| *** tongli has quit IRC | 22:29 | |
| *** trifon has quit IRC | 22:30 | |
| *** lpabon has quit IRC | 22:33 | |
| notmyname | TOO MUCH MAGIC!!! | 22:36 |
| gmmaha | notmyname: like peluse i am also getting error on functests with testr, but its varied errors on each run. 1st run - 68, 2nd run - 55, 3rd run 78, 4th run 44 :| | 22:37 |
| *** tsg has joined #openstack-swift | 22:43 | |
| mattoliverau | Swift: Where magic we don't understands happens. | 22:50 |
| mattoliverau | I was looking at the testr change yest arvo and having a few problems as well. I'm going to continue playing this morning. | 22:51 |
| mattoliverau | oh and hey gmmaha :) | 22:51 |
| gmmaha | hey mattoliverau :) | 22:52 |
| notmyname | mattoliverau: peluse: gmmaha: thank you for looking at those. unlike many other patches, that one has external time dependencies. the results of that work are what's given to the foundation board and what they approve for defcore stuff | 22:52 |
| gmmaha | ok, last week i thought i was crazy.. | 22:52 |
| gmmaha | thank god its not just me.. | 22:52 |
| notmyname | and we already missed the board meeting that happened in tokyo | 22:52 |
| *** trifon has joined #openstack-swift | 22:53 | |
| *** pdardeau has quit IRC | 22:53 | |
| mattoliverau | gmmaha: were all a little crazy really ;) | 22:53 |
| gmmaha | haha :D | 22:54 |
| notmyname | pay attention to this http://lists.openstack.org/pipermail/openstack-dev/2015-November/080294.html | 22:54 |
| mattoliverau | notmyname: kk, time to get out my square peg round hole bat :P | 22:54 |
| *** ChanServ changes topic to "Review Dashboard: https://goo.gl/iLt6Ld | Get an IRC bouncer at http://znc.in | Logs: http://eavesdrop.openstack.org/irclogs/%23openstack-swift/ | DocImpact changes http://lists.openstack.org/pipermail/openstack-dev/2015-November/080294.html" | 22:55 | |
| gmmaha | notmyname: thanks.. will let this go for now.. | 22:56 |
| notmyname | gmmaha: the testr failures? please don't let it go. I was saying that to encourage you to keep working on it ;-) | 22:57 |
| gmmaha | :D | 22:57 |
| gmmaha | since we missed the board meeting, i thought you were hinting towards letting it go until the next summit.. ;) | 22:58 |
| notmyname | I think there's another board meeting in the first half of december | 22:58 |
| notmyname | then at the next summit | 22:58 |
| gmmaha | ohh cool.. | 22:58 |
| notmyname | and with the way defcore works, there's basically a year lead time. so if our tests show new things that can be tested for, it will take a year to get them to be officially tested | 22:59 |
| tsg | zaitcev, we are discussing on -infra if liberasurecode package is available for F22 - I see it in "approved" status here: https://admin.fedoraproject.org/pkgdb/package/liberasurecode. Do you know when it would appear in F22 updates? | 22:59 |
| hogepodge | happy to answer any defcore questions people might have | 23:00 |
| zaitcev | tsg: I'm filing for F23 right not, and F22 has to follow it because version is the same. | 23:00 |
| *** km has joined #openstack-swift | 23:00 | |
| notmyname | hogepodge: just trying to get the testr changes in. getting weird failures in the gate | 23:00 |
| notmyname | hogepodge: when's the next board meeting? | 23:00 |
| *** km is now known as Guest31166 | 23:01 | |
| tsg | zaitcev, ok thanks .. I will change https://review.openstack.org/#/c/247936/ to install liberasurecode only >=f23 | 23:01 |
| hogepodge | December 3 | 23:01 |
| hogepodge | Conference call | 23:01 |
| notmyname | hogepodge: yikes. which means that any new capabilities discovered from the in-tree tests won't be ready by then | 23:02 |
| notmyname | hogepodge: but my understanding (from our conversation at the summit) is that new tests for existing capabilities can be added pretty easily. | 23:03 |
| *** david-lyle has joined #openstack-swift | 23:03 | |
| notmyname | hogepodge: does that require board action? or it's just written down? | 23:03 |
| hogepodge | notmyname: we should check with zehicle on when we want to set the new guidelines | 23:03 |
| notmyname | hogepodge: I'm much more concerned (short term) in adding new tests to the existing capabilities. i'm not as worried right now about adding new capabilities | 23:04 |
| *** david-lyle has quit IRC | 23:05 | |
| hogepodge | notmyname: totally understand, for some reason I think we have until January, but I want to make sure | 23:05 |
| notmyname | ok | 23:05 |
| notmyname | still, not much time | 23:05 |
| hogepodge | Just asked Rob to pop in and clarify. | 23:05 |
| notmyname | wait. have until january for what? | 23:05 |
| *** zehicle has joined #openstack-swift | 23:06 | |
| zehicle | o/ | 23:06 |
| hogepodge | get the swift in-tree tests added to the capabilities. | 23:06 |
| notmyname | ok | 23:06 |
| notmyname | that's better than "next week" what with thanksgiving this week | 23:06 |
| hogepodge | zehicle: what's our timing to get the swift in-tree tests added to existing capabilities? | 23:06 |
| hogepodge | zehicle: does it have to be ready by Dec 3 board meeting? | 23:06 |
| hogepodge | zehicle: I thought we had until January | 23:07 |
| tsg | zaitcev, sorry, may be I misunderstood - is F23 liberasurecode package already available, or it will be, after you "submit" it? :) | 23:09 |
| tsg | (zaitcev, I still see it under "testing") | 23:10 |
| zaitcev | tsg: it sits there for 3 days unless someone gives it more "karma" | 23:11 |
| zehicle | to existing caps? could happen for 2016.01 | 23:11 |
| zehicle | just need a patch for that | 23:11 |
| zehicle | hogepodge, no. | 23:11 |
| notmyname | zehicle: ok, so new tests to existing capabilities does not require a board action? just requires a patch to a json file? | 23:11 |
| zehicle | hogepodge, 2016.01 will likely really be 2016.02 (Feb meeting) | 23:11 |
| tsg | zaitcev, ok - do you have a tracker issue? | 23:12 |
| zehicle | notmyname, it can only be before the board approves | 23:12 |
| notmyname | zehicle: what does that mean? before the board approves what? | 23:12 |
| zaitcev | tsg: The Vote Formerly Known As Karma is now called "Feedback" in the UI in https://bodhi.fedoraproject.org/updates/FEDORA-2015-3889c3fcba | 23:12 |
| zehicle | once the Guideline is approved, we can't add tests | 23:12 |
| notmyname | zehicle: ok | 23:12 |
| zehicle | we cannot add new capabilities to the draft | 23:13 |
| notmyname | right | 23:13 |
| zehicle | tests are a level below caps | 23:13 |
| notmyname | zehicle: so what's the next guideline the board is approving? where are we now? | 23:13 |
| zehicle | so, new capabilities will have to wait until 2016.07 | 23:13 |
| notmyname | zehicle: (thinking timeframes) | 23:13 |
| notmyname | ok | 23:13 |
| zaitcev | tsg: until it's in stable, one can pull it from so-called "testing" repository, which is normally installed but not enabled by default. | 23:13 |
| zehicle | the capabilities for 2016.07 have to be in before the Austin summit | 23:13 |
| notmyname | zehicle: including all the scoring? | 23:14 |
| zehicle | notmyname, yes | 23:14 |
| tsg | zaitcev, got it. We can assume that the CI slaves won't have the repo enabled by default so it won't be available | 23:14 |
| notmyname | zehicle: and the tests? | 23:14 |
| zehicle | notmyname, ideally but there's some flexibility | 23:14 |
| notmyname | ideally what? | 23:15 |
| zehicle | ideally, we'd have all the tests for a Cap. Can get by w/ most of the tests. | 23:15 |
| zehicle | Capabilties without tests would not be ok | 23:15 |
| notmyname | what about tests for existing capabilities? | 23:15 |
| zehicle | those can come in now | 23:15 |
| notmyname | I'm not too worried about new capabilities yet | 23:15 |
| zaitcev | tsg: Probably a safe assumption. I checked and it's not on my laptop in F23. I guess it was split off into its own package or something. | 23:16 |
| zehicle | to the draft and future guidelines | 23:16 |
| notmyname | yeah, but the patch to get the tests integrated with tempest is ...struggling | 23:16 |
| zehicle | they cannot go into approved ones. we can only remove tests after approval | 23:16 |
| notmyname | so I'm trying to figure out what the rough time deadlines are | 23:16 |
| tsg | zaitcev, can I provide some level of testing and affect the "karma" in any way? | 23:16 |
| notmyname | zehicle: what's the next one that will be approved? | 23:16 |
| tsg | zaitcev, ie "feedback" | 23:16 |
| zehicle | notmyname, 2016.01 | 23:16 |
| notmyname | zehicle: it will be approved in january? or it is approved before? when should a patch be landed by? | 23:17 |
| zehicle | late January (maybe early Feb) | 23:17 |
| notmyname | ok, thanks | 23:17 |
| zehicle | the patch should land in early Jan to be safe | 23:17 |
| zehicle | (sooner is better) | 23:18 |
| notmyname | of course :-) | 23:18 |
| zaitcev | tsg: If you have a Fedora account, yes. | 23:18 |
| * zehicle DefCore is a tad distracted by the Oracle patch anyway | 23:18 | |
| notmyname | I'm pretty sure I don't want to know what that is ;-) | 23:18 |
| tsg | zaitcev, ok let me try | 23:18 |
| zehicle | notmyname, no, you probablly don't. has to do w/ hypervisors and zones | 23:18 |
| notmyname | and oracle ;-) | 23:19 |
| zehicle | and easily combustable materials | 23:19 |
| notmyname | hogepodge: zehicle: thanks for the defcore info | 23:20 |
| hogepodge | sure | 23:21 |
| notmyname | how does reading a config file not have a value but passing though ConfigParser() shows ky/val pairs? | 23:21 |
| notmyname | clayg: ^ | 23:22 |
| notmyname | Zyric1: still trying to figure out those config things on your patch. not sure yet what's going on. did you find anything? | 23:22 |
| *** david-lyle has joined #openstack-swift | 23:23 | |
| zaitcev | remote: FATAL: W refs/heads/f22 liberasurecode zaitcev DENIED by refs/heads/f[0-9][0-9] | 23:24 |
| notmyname | portante: I'm currently blaming you for in-proc func tests ;-) | 23:24 |
| zaitcev | tsg: oh sheesh, I guess it's F23 then | 23:24 |
| portante | notmyname: yes, you should, that is what I used to help find bugs in swift | 23:24 |
| portante | ;) | 23:24 |
| notmyname | lol | 23:24 |
| notmyname | portante: currently there's lots of magic and it's breaking the gate but not SAIO. so there | 23:25 |
| portante | breaking the gate ... oh, yes, now I remember, openstack gate jobs, those things ... ;) | 23:25 |
| notmyname | magic == I don't know how to explain what I'm seeing yet | 23:25 |
| portante | let me know, I can help debug if you'd like | 23:25 |
| *** david-lyle has quit IRC | 23:25 | |
| notmyname | portante: yeah, you're back in this "oh I can just push code and it lands right away" fairy-tale land ;-) | 23:26 |
| portante | ;) | 23:26 |
| zehicle | notmyname, glad to help! | 23:26 |
| tsg | zaitcev: thanks - not sure I understand why it was denied for F22 - freeze? .. let me see if I can other folks to test F23 and update the request | 23:27 |
| tsg | s/can other/can get other | 23:28 |
| zaitcev | tsg: Yeah, seems way early. The f21 would be denied because it went into maintenance. | 23:28 |
| tsg | zaitcev: ok .. let me make sure that the f21 gate jobs are in fact non-voting at this point and scheduled for removal end of this month | 23:28 |
| notmyname | ah ha! less magic now! | 23:29 |
| notmyname | thanks swifterdarrell for being my rubber duck | 23:29 |
| zaitcev | tsg: I'm movng on updating python-pyeclib in f23, so we have at least something complete before we deal with f22. | 23:29 |
| openstackgerrit | Ganesh Maharaj Mahalingam proposed openstack/swift: Unit tests for account/backend.py https://review.openstack.org/240613 | 23:30 |
| notmyname | Zyric1: progress :-) | 23:30 |
| *** kei_yama has joined #openstack-swift | 23:30 | |
| Zyric1 | notmyname: Sorry! No, not yet. I got a bit sidetracked with setting up my Outreachy blog, my bad. | 23:30 |
| tsg | zaitcev, ok .. hope you won't mind another update to pyeclib soon - we have several new updates on master and we are about to tag (after removing the bundled liberasurecode pkg) | 23:31 |
| zaitcev | tsg: Not at all, be my guest. | 23:31 |
| tsg | zaitcev: thanks! | 23:32 |
| zaitcev | tsg: Can I mix the upcoming PyECLib with liberasurede 1.1.0? | 23:32 |
| tsg | zaitcev: yes, in fact the upcoming PyECLib is also backward compatible with < 1.1.0 | 23:32 |
| tsg | zaitcev, its just that you need at least 1.1.0 for 'liberasurecode_rs_vand' if you want to use that for swift unit tests | 23:33 |
| zaitcev | tsg: It would be great if you and Kevin allowed for a window of 1 release back and forth, e.g. pyeclib 1.1.1 working with liberasurecode 1.1.0. Of course not with 1.0.9, that would be too much to ask. | 23:33 |
| zaitcev | tsg: I see, thanks. | 23:33 |
| *** david-lyle has joined #openstack-swift | 23:35 | |
| notmyname | ah ha! found it | 23:36 |
| notmyname | Zyric1: ^ | 23:36 |
| notmyname | portante: ^ found the problem. still not particularly happy with you ;-) | 23:36 |
| *** trifon has quit IRC | 23:36 | |
| notmyname | turns out that setting up rings also happens to helpfully rewrite the config file. found with judicious printf debugging | 23:37 |
| Zyric1 | notmyname: you're my hero! :) | 23:38 |
| notmyname | Zyric1: well, not quite. it's a problem with your code ;-) | 23:38 |
| portante | notmyname: sorry. :( | 23:39 |
| notmyname | Zyric1: but now I know where to look, so the fix should be easier :-) | 23:39 |
| portante | I am not causing problems in another environment, if that is any consolation. :) | 23:39 |
| notmyname | portante: no worries. think of it as "you're not forgotten" ;-) | 23:39 |
| *** sgundur has quit IRC | 23:40 | |
| notmyname | portante: I'd rather have you back causing problems in swift :-) | 23:40 |
| Zyric1 | notmyname: while it may be my code that's the problem, I know my code a lot better than I understand rings and much of swift at the moment so your help with the debugging is super helpful | 23:40 |
| portante | thanks! | 23:40 |
| notmyname | Zyric1: "your code as the problem" is a little strong :-) | 23:41 |
| *** david-lyle has quit IRC | 23:41 | |
| *** david-ly_ has joined #openstack-swift | 23:41 | |
| notmyname | Zyric1: there's an interaction we didn't see at first, and that's what's causing problems | 23:41 |
| notmyname | Zyric1: here's what's happening | 23:41 |
| *** sgundur has joined #openstack-swift | 23:41 | |
| notmyname | Zyric1: the tox -e func job runs what's called in-process tests. (written originally by portante, hence my bugging him [in jest]) | 23:42 |
| notmyname | Zyric1: those are designed to let you run functional tests without having an external swift install | 23:42 |
| notmyname | Zyric1: to do that, it sets up some defaults in configs, based on the sample config files, and starts up some processes | 23:43 |
| notmyname | Zyric1: however, when it's setting up a swift.conf file (which includes the storage policy definitions), it calls get_info() on at least the first storage policy | 23:43 |
| notmyname | that first one defaults to "Policy-0" as the name (for historic reasons) | 23:44 |
| notmyname | Zyric1: so your patch adds in the aliases, and the swift.conf test setup iterates over what that responds with and writes it in a config | 23:44 |
| notmyname | Zyric1: the aliases attribute referred to in teh config mapping is actually stored as a list, not a string. so when it's coerced to a string, you get the "['Policy-0']" form | 23:45 |
| notmyname | Zyric1: and that "blows up", as they say | 23:45 |
| notmyname | Zyric1: IMO here's how you should fix it. look at the "default" and "deprecated" values in the config map. they refer to a method that gives the right value, so you don't get the string coercion on the bools | 23:47 |
| notmyname | Zyric1: I think this patch needs something similar. eg map aliases to "aliases_readable_list" or something | 23:48 |
| notmyname | Zyric1: does that make sense? | 23:48 |
| Zyric1 | notmyname: Yes, that makes much more sense! Thank you for the clear explanation :) | 23:49 |
| notmyname | Zyric1: I'm writing a diff | 23:51 |
| *** sgundur has quit IRC | 23:52 | |
| notmyname | Zyric1: also, just as a code style note, we've got version control for the code, so there's no reason to keep commented out code in the file | 23:52 |
| *** sgundur has joined #openstack-swift | 23:53 | |
| *** ho has joined #openstack-swift | 23:54 | |
| ho | good morning! | 23:55 |
| mattoliverau | ho: morning | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!