| *** gyee has quit IRC | 00:01 | |
| *** mwstorer has quit IRC | 00:01 | |
| clayg | creiht: peluse_: notmyname: so on the case insenitive names, it seems to make the most sense to use the name as it appears in swift.conf verbatium in responses, and only apply "insensitivity" to client requests and handling duplicate names? | 00:04 |
|---|---|---|
| clayg | ok, so hopefully liberal in what we except and strict in what we respond applies | 00:08 |
| notmyname | accept | 00:08 |
| clayg | the /info response will include *exactly* what you put in swift.conf | 00:08 |
| notmyname | to confim in IRC what I said in the office.. | 00:09 |
| notmyname | clayg: +1 that sounds right | 00:09 |
| clayg | when you get a policy name in a header it's title case normalized | 00:09 |
| peluse_ | ditto | 00:14 |
| clayg | peluse_: currently discussing with notmyname if _ should be valid in a policy name | 00:14 |
| clayg | i added one and the response from account is X-Account-Storage-Policy-Gold_Stuff-Bytes-Used | 00:14 |
| notmyname | do we allow it in metadata keys now? | 00:15 |
| peluse_ | kinda wierd | 00:15 |
| clayg | which swift/eventlet seem fine with - but i don't know if the header would be dropped if you stuck some weird http terminating/load-balancing device in the mix | 00:15 |
| notmyname | https://gist.github.com/notmyname/2b2b1e7c361a729927b4 | 00:16 |
| notmyname | so we do weird stuff | 00:16 |
| notmyname | set a metadata key with an underscore and we change it to a dash | 00:16 |
| clayg | i think again the strict response may tell us that we should convert _ to - in the account response, and also enfoce you can't have Policy_0 and policy-0? | 00:16 |
| notmyname | clayg: ie normalization is .replace('-','_').upper() | 00:17 |
| clayg | probably - did you see where the swift is doing that for the metadata keys? | 00:17 |
| notmyname | clayg: no, not in the code yet. | 00:18 |
| *** zhiyan_ is now known as zhiyan | 00:20 | |
| *** judd7 has joined #openstack-swift | 00:20 | |
| zaitcev | "a daemon ... works tirelessly to identify and rectify this potential scenario." | 00:23 |
| clayg | zaitcev: too much prose in there for you? | 00:23 |
| clayg | zaitcev: or anthropomorphism | 00:24 |
| zaitcev | clayg: I think the prose is not appropriate in a document of such kind. | 00:24 |
| notmyname | clayg: swob does it | 00:25 |
| notmyname | changes _ to - | 00:25 |
| clayg | notmyname: well then why doesn't it do it in header response from account server when I have a storage policy name with a _? | 00:25 |
| clayg | I'm feeling more and more like I should force the values in swift.conf to be titlecase and - only? | 00:26 |
| notmyname | well it's not super straightforward... | 00:26 |
| *** kashyapk has joined #openstack-swift | 00:27 | |
| notmyname | clayg: looks like any swob.Request.headers.keys() will replace _ with - | 00:28 |
| *** zhiyan is now known as zhiyan_ | 00:29 | |
| *** dmorita has joined #openstack-swift | 00:30 | |
| *** matsuhashi has joined #openstack-swift | 00:30 | |
| *** kashyapk has quit IRC | 00:32 | |
| peluse_ | clayg: I agree with you, I don't think changing the name of a policy on someone is a good idea | 00:33 |
| peluse_ | I'd be like "WTF?" | 00:33 |
| notmyname | bah. swob.Request.headers = HeaderEnvironProxy which does the translation. swob.Response.headers = HeaderKeyDict which only does the title case | 00:34 |
| clayg | notmyname: so it's fine I can normalize the header in the account server response (which is the only place we stuff the name into a header *key*) | 00:35 |
| notmyname | clayg: peluse_: so if policy.name = "thisIs_a_GOLD_policy" you'd error? | 00:35 |
| clayg | notmyname: maybe, or just make it Thisis-A-Gold-Policy and call it a day | 00:35 |
| clayg | people who like to use CamelCase are going to be mad though :\ | 00:35 |
| notmyname | clayg: and make it that way everywhere. no matter what's in swift.conf | 00:36 |
| clayg | notmyname: somethign like that is one option yeah | 00:36 |
| notmyname | for a second choice, that's ok. the consistency is important I think | 00:36 |
| peluse_ | you can please some of the people all of the time, all of the people some of the time... but | 00:36 |
| * peluse_ heads out for some grub,,, back later | 00:38 | |
| notmyname | peluse_: and all of the people all of the time, if you are a dictator | 00:38 |
| peluse_ | :) | 00:38 |
| *** ozialien has quit IRC | 00:42 | |
| *** shri has quit IRC | 00:47 | |
| *** Edward-Zhang has quit IRC | 00:48 | |
| *** zhiyan_ is now known as zhiyan | 00:50 | |
| clayg | sweet underscores are out - for reasons | 00:54 |
| *** zhiyan is now known as zhiyan_ | 00:59 | |
| *** zhiyan_ is now known as zhiyan | 01:09 | |
| *** ujjain has quit IRC | 01:15 | |
| *** swat30 has quit IRC | 01:15 | |
| *** Thurloat has quit IRC | 01:16 | |
| creiht | clayg: agreed | 01:19 |
| clayg | so i guess since we accept spaces and utf-8 in metadata keys they're ok in policy names | 01:19 |
| *** ujjain has joined #openstack-swift | 01:19 | |
| creiht | lol | 01:22 |
| creiht | that doesn't mean we have to support that | 01:22 |
| creiht | utf-8 policy names seem problematic | 01:22 |
| creiht | maybe we should only support uuids :) | 01:22 |
| clayg | creiht: yeah idk, someone is going to have a trademark on something and want that to be their storage policy name - i don't know if it's a good idea, but I don't how much we need to hold peoples hands either? | 01:23 |
| creiht | hah | 01:24 |
| zaitcev | How do you post spaces into X-Storage-Policy-A space B-Object-Count: | 01:26 |
| clayg | zaitcev: same way you do for X-Account-Meta-Something something: Foo ??? | 01:27 |
| clayg | zaitcev: anyway, the name of the storage policy is normally a header *value* - but we stick it into the header keys of the account response for per policy object/byte counts - well... i guess you saw that | 01:28 |
| clayg | zaitcev: but you don't POST those values anymore than you do X-Account-Object-Count | 01:28 |
| clayg | er... POST those keys | 01:28 |
| clayg | heh /info response is json so utf comes back unicode literal yay! | 01:29 |
| creiht | lol | 01:29 |
| *** diegows has quit IRC | 01:30 | |
| clayg | ipyt | 01:31 |
| clayg | hrmm... so ConfigParser returns encoded byte array and in python3 it's just a "str" (which are always unicode i guess) | 01:37 |
| *** Edward-Zhang has joined #openstack-swift | 01:42 | |
| *** nosnos has joined #openstack-swift | 01:47 | |
| clayg | so I think the most defensible posistion is to be overly restrictive with a nod that it could more easily be expanded than contracted | 01:57 |
| *** swat30 has joined #openstack-swift | 02:00 | |
| *** NM has joined #openstack-swift | 02:01 | |
| *** nthacker has joined #openstack-swift | 02:02 | |
| nthacker | hi all, so i noticed some bugs have patches in them, but they are still open....the patch looks like they fix the issues | 02:03 |
| nthacker | so im curious if the fix can be pushed by the patch owner? or someone can just test it and push it on behalf of the patch author? | 02:03 |
| nthacker | jus curious here.... | 02:03 |
| nthacker | since i was looking to fix some bugs myself :P | 02:03 |
| zaitcev | ask the guy who worked on the patch, and if he agrees or not responsive, submit it to Gerrit. | 02:04 |
| *** NM has quit IRC | 02:04 | |
| nthacker | zaitcev: Ok, thanks - ask him where? here? or email? | 02:04 |
| clayg | nthacker: i think sometimes the bugs with patches don't really include test cases, or cover all corner cases - so it'd be great to get something up in a review - but it make take some additional work beyond just submitting the patch file | 02:05 |
| clayg | but in either case no, I don't think you need to wait on the bug author to submit the patch - or even contact them - if you put the "fixes #<bug-number>" in the commit I think bots are supposed to comment on launchpad | 02:05 |
| nthacker | clayg: Ok. that makes sense. actually i think it was your patches, but I'll try out the patch and verify testing with whatever cases i can come up with | 02:06 |
| nthacker | clayg: dont all checkins need to be reviewed in gerritt? | 02:06 |
| clayg | yeah... there's a "submitting to openstack" page out there on the internet somewhere | 02:07 |
| *** swat30 has quit IRC | 02:08 | |
| clayg | peluse_: so I acctually like Policy-0 now... it's like idk... the beginning of everything, or Unimatrix Zero or something | 02:09 |
| clayg | I couldn't see it somehow when it was Policy_0 | 02:10 |
| *** ozialien has joined #openstack-swift | 02:12 | |
| mattoliverau | nthacker: if you are submitting someone else's patch and/or adding to someone's patch then add a 'co-authored-by'. So the orginal author is referenced. | 02:14 |
| clayg | mattoliverau: he's taking a .patch file attached to an lp bug and trying to get it into gerrit | 02:15 |
| mattoliverau | Fair enough, but did he write the code? if not, when we submits it, it'll be in his name, and unless he has the original authors aporoval there can be legal ramificaitons.. I only know this because someone spoke up about a similar issue (complained) to one of the cores in the infra channel a while ago (pre-summit). The summit also had discussions about it (in the CLA metting). The CLA states that any | 02:21 |
| mattoliverau | code you submit you have the right to. (in fancier terms of course :P). So by adding the co-auther bit gives them reconition for their work. | 02:21 |
| clayg | i think if you attach a .patch to a bug report you're giving up copy-right - otherwise you should inclue a license file with the .patch | 02:22 |
| clayg | but INAL | 02:22 |
| mattoliverau | clayg: very true | 02:23 |
| clayg | it's probably more complicated if you've signed the CLA and submitted the patch to gerrit and then someone wants to take over that change - worse if they remove any CopyRight lines you added to the change before it got approved/merged | 02:24 |
| mattoliverau | clayg: that was closer to the situation that happened a few months ago. Someone took the contents of anoher gerrit patch, added to it and submitted it as there work. The original change might have also been abandoned, I can't remember. So maybe LP patch is different. But still I think it's better safe then sorry :) | 02:26 |
| mattoliverau | And hey it's only 1 line at the end of the commit message :) | 02:26 |
| nthacker | mattoliverau: ok thanks - yes i actually agree with that style | 02:34 |
| gholt | Man, you guys keep thinking of great stuff I'm not catching. Case insensitive policy names, don't overly lazy load rings (if at all), etc. | 02:42 |
| clayg | creiht's good at the api stuff | 02:43 |
| clayg | dfg's good at the not be horribly broken stuff | 02:43 |
| gholt | Heheh | 02:43 |
| clayg | go team swift! | 02:43 |
| zaitcev | I'm good at nitpicking | 02:43 |
| zaitcev | *puppyeyes* | 02:44 |
| gholt | I don't know about the schema thing. I think all the schema stuff has been in the backend area from the start so it just seemed logical to be there to me. | 02:44 |
| gholt | Haha, and zaitcev chimes in right when I'm talking about one of his complaints. :) | 02:44 |
| clayg | gholt: did I leak some schema stuff out of backend? :\ | 02:45 |
| *** matsuhashi has quit IRC | 02:45 | |
| gholt | Well, earlier today you said you were going to fart types; but you weren't really in the mood for that 5-year-old humor. ;D | 02:45 |
| creiht | hehe | 02:46 |
| zaitcev | clayg: I haven't seen the code yet. I just think it was underdocumented. | 02:46 |
| clayg | i can't tell if you guys are messing with me | 02:46 |
| gholt | Assume always. | 02:47 |
| *** haomaiwa_ has joined #openstack-swift | 02:47 | |
| creiht | lol | 02:47 |
| *** swat30 has joined #openstack-swift | 02:48 | |
| zaitcev | except me. I never mess with people... in English | 02:48 |
| zaitcev | I can place a play-doh on your chair though | 02:48 |
| clayg | i don't know what to say... i'm just trying to keep high spirts while we noodle around with docs and config file parsing - cause the really good stuff is yet to come | 02:48 |
| clayg | lol | 02:48 |
| creiht | clayg: yeah we are getting there | 02:48 |
| gholt | zaitcev: :) Yeah... right. Of course, you might not think you're missing with people; it's just your way. | 02:48 |
| creiht | takes some time to digest everything | 02:48 |
| creiht | that sounds like a book | 02:49 |
| creiht | "The Way of Zaitcev" | 02:49 |
| clayg | yeah... honestly I've always felt it takes longer to review code than to write it - and this crap took like what? a year? we're screwed | 02:49 |
| creiht | lol | 02:49 |
| clayg | you all are doing great though thanks for everyone coming together... gholt's right we're still gunna be at in CO - but then you guys can throw stuff at me | 02:50 |
| clayg | so the more we get done tomorrow the better i suppose | 02:50 |
| clayg | ah crap... i was gunna get the policy_string munge on q/tmp dirs in with tonights rebase | 02:50 |
| zaitcev | I thought we ought to complete SP reviews before Longmont and focus on EC there. | 02:50 |
| clayg | stupid unicode | 02:50 |
| zaitcev | I know I'm behind on this | 02:50 |
| clayg | zaitcev: i'd like to mostly just talk about PBE next week :D | 02:51 |
| clayg | s/talk/work-on/ | 02:51 |
| creiht | PBE? | 02:51 |
| zaitcev | clayg: Constantin poked me about some changes in PBE. He wants some method protocols changed to accomodate ZeroVM better. | 02:51 |
| creiht | oh yeah | 02:51 |
| clayg | zaitcev: oh ok... good! | 02:52 |
| creiht | clayg: I was thinking maybe we should spend one day doing a policies code roast | 02:52 |
| creiht | or half day | 02:52 |
| creiht | whatever it takes | 02:52 |
| clayg | creiht: i don't even know if that will be enough... | 02:54 |
| creiht | heh | 02:54 |
| *** haomaiwa_ has quit IRC | 02:55 | |
| *** haomaiwang has joined #openstack-swift | 02:55 | |
| clayg | gholt: zaitcev: oh you guys were asking about the thing where the auditor will automatically upgrade legacy schemas as it finds them instead of leaving them unmigrated forever until a write forces it? | 02:56 |
| zaitcev | clayg: I was | 02:56 |
| clayg | gholt: zaitcev: I'm pretty sure we want to force it... the container_sync schema change can like sit around forever (if you're not running the container-sync daemon) - and every get_info on those db's has to get the sqlite exception, trap it and retry the query. | 02:57 |
| zaitcev | clayg: I don't see a problem with it, and docs mention it in passing right at the end, but I thought it needed some kind of acknowledgement, like the way the new table for account was. | 02:57 |
| *** shakayumi has quit IRC | 02:58 | |
| clayg | oh... yeah... maybe... probably something updated on the page that talks about how the auditors work - oh wait... where is that page? | 02:58 |
| zaitcev | clayg: my plan was actually consider those notes I left as margin marks and get back to them when I saw the actual code | 02:58 |
| clayg | zaitcev: we'll they'll get burried in tonights rebase if I don't address them somehow | 02:59 |
| clayg | zaitcev: do you want a sentence that says something like "Policies were not implemented at the account level because of the overhead of adding support to all legacy auth systems"? | 03:03 |
| *** nthacker has quit IRC | 03:03 | |
| zaitcev | clayg: Something like that. Because I thought they were not because of existing users who have single sign-on or other and need access to 2 policies while only having 1 account. | 03:04 |
| clayg | yeah basically - but that's sort of a specific subset | 03:04 |
| clayg | zaitcev: what does "one -s" mean | 03:08 |
| zaitcev | things happens | 03:08 |
| zaitcev | Is "Storage Policies" singular or plural? | 03:08 |
| clayg | i.... don't know? | 03:09 |
| clayg | s/fits/fit/ maybe? | 03:09 |
| *** ozialien has quit IRC | 03:10 | |
| *** haomaiw__ has joined #openstack-swift | 03:12 | |
| *** haomaiwang has quit IRC | 03:13 | |
| clayg | zaitcev: re downgrade - it won't work | 03:25 |
| zaitcev | clayg: so... doc change? | 03:26 |
| zaitcev | personally I'm okay with no downgrade, although operators may be not as enthusiastic, in case they hit any regressions | 03:26 |
| clayg | zaitcev: yeah well i'm just gunna say "don't" for now... i guess the change log will need something a bit more scary or... idk... maybe i'll have to do something in the triggers... but it's gunna be slower :\ | 03:27 |
| *** haomaiwa_ has joined #openstack-swift | 03:27 | |
| clayg | damn | 03:27 |
| *** ppai has joined #openstack-swift | 03:28 | |
| clayg | ok ok ok, it's not that bad... maybe | 03:28 |
| clayg | we could test it | 03:28 |
| clayg | i mean the old code maybe won't report bytes and counts correctly - but the triggers should still be doing the right thing | 03:30 |
| *** mrsnivvel has joined #openstack-swift | 03:30 | |
| *** haomaiw__ has quit IRC | 03:30 | |
| zaitcev | well we can just cross ourselves and jump into it, and deployers usually do not downgrade anyway | 03:36 |
| zaitcev | we need to be sensitive to small stuff like that DB cursor thing | 03:36 |
| clayg | I don't really have a good answer | 03:39 |
| *** nosnos has quit IRC | 03:40 | |
| clayg | zaitcev: I think i was able to address most of your comments except why not a new config file... frankly that decsion was made before I started working on policies - so you might have to ask peluse_ or torgomatic | 03:43 |
| clayg | zaitcev: I think we could change it, but I don't really see it as necessary | 03:43 |
| *** matsuhashi has joined #openstack-swift | 03:47 | |
| *** matsuhashi has quit IRC | 03:54 | |
| *** matsuhashi has joined #openstack-swift | 03:54 | |
| *** matsuhashi has quit IRC | 03:57 | |
| *** matsuhashi has joined #openstack-swift | 03:58 | |
| *** matsuhashi has quit IRC | 04:03 | |
| *** matsuhashi has joined #openstack-swift | 04:03 | |
| *** zaitcev has quit IRC | 04:17 | |
| *** psharma has joined #openstack-swift | 04:22 | |
| *** kashyapk has joined #openstack-swift | 04:33 | |
| *** nosnos has joined #openstack-swift | 04:42 | |
| *** swat30 has quit IRC | 05:01 | |
| *** zhiyan is now known as zhiyan_ | 05:13 | |
| *** zhiyan_ is now known as zhiyan | 05:35 | |
| *** zhiyan is now known as zhiyan_ | 05:47 | |
| *** ZBhatti has quit IRC | 05:59 | |
| *** zhiyan_ is now known as zhiyan | 06:03 | |
| *** saurabh_ has quit IRC | 06:10 | |
| *** zhiyan is now known as zhiyan_ | 06:19 | |
| *** zhiyan_ is now known as zhiyan | 06:20 | |
| *** ujjain has quit IRC | 06:21 | |
| *** ujjain has joined #openstack-swift | 06:21 | |
| *** matsuhashi has quit IRC | 06:30 | |
| *** matsuhashi has joined #openstack-swift | 06:31 | |
| mattoliverau | well im going to call it a day, and a week! have a great weekend all! | 07:04 |
| *** zhiyan is now known as zhiyan_ | 07:06 | |
| *** matsuhashi has quit IRC | 07:11 | |
| *** matsuhas_ has joined #openstack-swift | 07:16 | |
| *** matsuhas_ has quit IRC | 07:16 | |
| *** zhiyan_ is now known as zhiyan | 07:17 | |
| *** matsuhashi has joined #openstack-swift | 07:17 | |
| *** sungju has quit IRC | 07:28 | |
| *** nshaikh has joined #openstack-swift | 07:43 | |
| *** haomaiwa_ has quit IRC | 07:45 | |
| *** haomaiwang has joined #openstack-swift | 07:46 | |
| *** haomai___ has joined #openstack-swift | 07:49 | |
| *** haomaiwang has quit IRC | 07:53 | |
| *** nacim has joined #openstack-swift | 07:55 | |
| *** scotticus has quit IRC | 07:56 | |
| *** scotticus has joined #openstack-swift | 08:02 | |
| *** Midnightmyth has joined #openstack-swift | 08:18 | |
| *** bvandenh has joined #openstack-swift | 08:19 | |
| *** zhiyan is now known as zhiyan_ | 08:25 | |
| *** zhiyan_ is now known as zhiyan | 08:47 | |
| *** zhiyan is now known as zhiyan_ | 08:49 | |
| *** haomai___ has quit IRC | 08:52 | |
| *** haomaiwang has joined #openstack-swift | 08:53 | |
| *** Edward-Zhang has quit IRC | 09:03 | |
| *** nshaikh has quit IRC | 09:11 | |
| *** haomaiwang has quit IRC | 09:18 | |
| *** haomaiwang has joined #openstack-swift | 09:19 | |
| *** mkollaro has joined #openstack-swift | 09:22 | |
| *** nshaikh has joined #openstack-swift | 09:28 | |
| *** Midnightmyth has quit IRC | 09:35 | |
| *** mrsnivvel has quit IRC | 09:37 | |
| *** Midnightmyth has joined #openstack-swift | 09:42 | |
| *** mrsnivvel has joined #openstack-swift | 09:48 | |
| *** zhiyan_ is now known as zhiyan | 09:49 | |
| *** dmorita has quit IRC | 09:55 | |
| *** kashyapk has quit IRC | 09:58 | |
| *** zhiyan is now known as zhiyan_ | 09:59 | |
| *** kashyapk has joined #openstack-swift | 09:59 | |
| *** hxgqh1987 has joined #openstack-swift | 10:07 | |
| *** hxgqh1987 has quit IRC | 10:07 | |
| *** matsuhashi has quit IRC | 10:16 | |
| *** matsuhashi has joined #openstack-swift | 10:21 | |
| *** matsuhashi has quit IRC | 10:27 | |
| *** matsuhashi has joined #openstack-swift | 10:28 | |
| *** haomaiwang has quit IRC | 10:32 | |
| *** haomaiwa_ has joined #openstack-swift | 10:33 | |
| *** haomaiw__ has joined #openstack-swift | 10:35 | |
| *** haomaiwa_ has quit IRC | 10:39 | |
| *** zhiyan_ is now known as zhiyan | 10:50 | |
| *** zhiyan is now known as zhiyan_ | 10:59 | |
| *** matsuhashi has quit IRC | 11:00 | |
| *** matsuhashi has joined #openstack-swift | 11:00 | |
| *** kashyapk has quit IRC | 11:01 | |
| *** kashyapk has joined #openstack-swift | 11:02 | |
| *** NM has joined #openstack-swift | 11:02 | |
| *** kashyapk has quit IRC | 11:03 | |
| *** kashyapk has joined #openstack-swift | 11:04 | |
| *** matsuhashi has quit IRC | 11:05 | |
| *** kashyapk has quit IRC | 11:06 | |
| *** kashyapk has joined #openstack-swift | 11:07 | |
| *** kashyapk_ has joined #openstack-swift | 11:08 | |
| *** NM has quit IRC | 11:08 | |
| *** matsuhashi has joined #openstack-swift | 11:09 | |
| *** Midnightmyth has quit IRC | 11:11 | |
| *** kashyapk has quit IRC | 11:12 | |
| *** openstackgerrit has joined #openstack-swift | 11:18 | |
| *** openstackgerrit has quit IRC | 11:19 | |
| *** openstackgerrit has joined #openstack-swift | 11:24 | |
| *** openstackgerrit has quit IRC | 11:25 | |
| *** openstackgerrit has joined #openstack-swift | 11:25 | |
| *** mlipchuk has joined #openstack-swift | 11:30 | |
| *** mlipchuk has quit IRC | 11:48 | |
| *** zhiyan_ is now known as zhiyan | 11:51 | |
| *** swat30 has joined #openstack-swift | 11:52 | |
| *** haomaiw__ has quit IRC | 11:52 | |
| *** matsuhashi has quit IRC | 12:00 | |
| *** zhiyan is now known as zhiyan_ | 12:00 | |
| *** matsuhashi has joined #openstack-swift | 12:00 | |
| *** Midnightmyth has joined #openstack-swift | 12:01 | |
| *** matsuhashi has quit IRC | 12:02 | |
| *** matsuhashi has joined #openstack-swift | 12:02 | |
| *** bvandenh has quit IRC | 12:03 | |
| *** diegows has joined #openstack-swift | 12:05 | |
| *** matsuhashi has quit IRC | 12:12 | |
| *** matsuhashi has joined #openstack-swift | 12:13 | |
| *** matsuhashi has quit IRC | 12:13 | |
| *** openstackgerrit has quit IRC | 12:22 | |
| *** openstackgerrit has joined #openstack-swift | 12:22 | |
| *** Thurloat has joined #openstack-swift | 12:26 | |
| *** lpabon has joined #openstack-swift | 12:31 | |
| *** lpabon has quit IRC | 12:31 | |
| *** ppai has quit IRC | 12:37 | |
| gholt | Heh, "deleting the policy and associated rings would be problematic for existing data." Or maybe nearly catastrophic. | 12:38 |
| *** nosnos has quit IRC | 12:40 | |
| *** haomaiwang has joined #openstack-swift | 12:46 | |
| *** zhiyan_ is now known as zhiyan | 12:51 | |
| *** zhiyan is now known as zhiyan_ | 13:03 | |
| gholt | clayg: notmyname: Oh man, reviewing updates is going to be difficult as hell. ;) It was easy for the top level patchset, but the dependent ones... Can't just diff the previous patchset to the current since that contains everything from the dependency too. Hmm. | 13:04 |
| *** psharma has quit IRC | 13:08 | |
| gholt | I have a solution, albeit cumbersome. In this example, I'll be reviewing the changes to 96027 from its previous patchset 3: | 13:14 |
| gholt | First, I check out the current patchset and name that branch "rev96027". | 13:15 |
| gholt | Then I check out the dependency and name that branch "rev96026". | 13:15 |
| gholt | Next I check out the previous patchset 3 of 96027 and name that "anchor". | 13:15 |
| gholt | Now I do a "git log" to figure out the commit hash for the previous 96026 dependency in that particular version of history and "git revert" it. | 13:15 |
| gholt | Next up, I "git merge rev96026" into that branch (still on "anchor" at this point). | 13:15 |
| gholt | Now I can "git checkout rev96027" and do a "git diff anchor" and see the changes, yay! | 13:15 |
| *** kashyapk_ has quit IRC | 13:17 | |
| *** kashyapk has joined #openstack-swift | 13:17 | |
| gholt | I'm reminded of the saying, "If it was a pain in the ass to write, it should be a pain in the ass to read!" | 13:19 |
| *** NM has joined #openstack-swift | 13:22 | |
| *** kenhui has joined #openstack-swift | 13:26 | |
| *** judd7 has quit IRC | 13:27 | |
| *** mrsnivvel has quit IRC | 13:28 | |
| *** haomaiwang has quit IRC | 13:29 | |
| *** haomaiwa_ has joined #openstack-swift | 13:29 | |
| *** haomaiw__ has joined #openstack-swift | 13:31 | |
| *** Diddi has quit IRC | 13:35 | |
| *** haomaiwa_ has quit IRC | 13:35 | |
| gholt | Damn, this won't work well paste the second item in the chain though. I'm on the third now and it's... much worse. | 13:35 |
| *** tdasilva has joined #openstack-swift | 13:37 | |
| *** kashyapk has quit IRC | 13:37 | |
| gholt | Have to revert two commits from the old patchchain and merge in the two new commits from the new patchchain, but there are conflicts trying to the reverts and probably would be conflicts merging in the new stuff. It'll almost be impossible to review the diffs from each patchset individually unless someone comes up with something amazing I'm not thinking of. | 13:37 |
| *** kashyapk has joined #openstack-swift | 13:37 | |
| gholt | I'll just pause for a while and see what folks think of as it's currently nearly impossible for me to tell what changed from patch to patch in 96208 for example. | 13:39 |
| peluse_ | we'll talk more at the hackathon but now seems like a good time to plug the strategy of not using an integration branch for the EC work... straight to master with the EC policy disabled until complete... | 13:40 |
| gholt | I might be able to diff diffs and work from there, though that might also be a mess. | 13:44 |
| acoles | gholt: had the same sinking feeling myself this morning | 13:47 |
| acoles | gholt: but how about this... | 13:47 |
| acoles | create branch from master, then use the cherry pick command from gerrit for e.g. 96027/3 | 13:48 |
| acoles | create another branch_2 from master, cherry pick 96027/4 | 13:49 |
| acoles | on branch2, diff branch1 | 13:49 |
| acoles | i *think* i see only the relative changes on 96027 | 13:50 |
| gholt | I'm not sure you can cherry pick for things that truly depend on the prior commits though right? | 13:50 |
| acoles | gholt: i'm not expecting tests to pass, just hoping to see what changed between versions of that patch | 13:54 |
| *** zhiyan_ is now known as zhiyan | 13:54 | |
| gholt | Yeah, I'm just not sure the cherrypick would even succeed though. | 13:55 |
| gholt | The diff of diffs seems to work reasonably okay. | 13:56 |
| acoles | hmm, maybe further up the chain it will fall down | 13:56 |
| acoles | cherrypick i mean | 13:56 |
| *** jergerber has joined #openstack-swift | 13:56 | |
| gholt | For example, the diff of diffs shows, I think ;), that nothing changed specifically for 96028. | 13:56 |
| acoles | gholt: put a comment on that one then :) | 13:57 |
| *** haomaiw__ has quit IRC | 14:03 | |
| *** haomaiwa_ has joined #openstack-swift | 14:03 | |
| *** zhiyan is now known as zhiyan_ | 14:03 | |
| gholt | Damn, my diff of diffs seems to have fallen down at 96030; could just be user error though, which is quite likely for this messy sort of thing. | 14:06 |
| gholt | I guess I could just only work from the end of the chain, but then why bother having clayg split it all up. Even though it's a bit harder to really get going on the reviewing, it's probably just better to have one big commit. | 14:08 |
| tdasilva | gholt and others: if I run a "review -d" of the last commit, will that pull all the other commits? is that how are you guys testing this? | 14:10 |
| gholt | tdasilva: Yes | 14:10 |
| gholt | git review -d 96049 | 14:11 |
| tdasilva | gholt: ok, thanks | 14:11 |
| gholt | acoles: Oh, and it's made worse in that some of these have multiple patchsets per day as clayg and peluse_ work through things. | 14:12 |
| gholt | For example, it seems like there are zero changes in the last patchset for 96049. | 14:15 |
| gholt | That doesn't seem right, but maybe, just maybe, only a commit message somewhere changed on that one? | 14:16 |
| *** judd7 has joined #openstack-swift | 14:20 | |
| *** haomaiw__ has joined #openstack-swift | 14:21 | |
| gholt | Poor guys. Someone had to be first to walk us through this sort of thing. Heh | 14:22 |
| *** shakamunyi has joined #openstack-swift | 14:22 | |
| *** haomaiwa_ has quit IRC | 14:23 | |
| acoles | gholt: we need another button in gerrit :) | 14:24 |
| creiht | is anyone going to the hackathon flying back on friday? | 14:27 |
| *** openstackgerrit has quit IRC | 14:36 | |
| *** openstackgerrit has joined #openstack-swift | 14:36 | |
| peluse_ | creiht: I'm flying home Thu eve | 14:43 |
| creiht | I have to fly back on Friday and wanted to see if anyone might be able to give me a ride :) | 14:44 |
| *** kashyapk has quit IRC | 14:46 | |
| *** kashyapk has joined #openstack-swift | 14:47 | |
| portante | creiht: sorry, flying back Thu night red eye | 14:47 |
| peluse_ | I'm getting in Sun afternoon and getting a few things ready in the office if anyone is in early enough for dinner at 6ish or 7 | 14:49 |
| *** kenhui has quit IRC | 14:49 | |
| *** kenhui has joined #openstack-swift | 14:49 | |
| *** kashyapk has quit IRC | 14:51 | |
| openstackgerrit | paul luse proposed a change to openstack/swift: Make quarantine and tmp dirs in DiskFile Storage Policy Aware https://review.openstack.org/96313 | 14:51 |
| portante | peluse_: coming in at 10:15 PM to Denver, sorry | 14:52 |
| *** zhiyan_ is now known as zhiyan | 14:55 | |
| acoles | peluse_: arrive sun pm too late for dinner sorry creiht: leave late thurs | 14:56 |
| creiht | no worries, I think I have it figured out | 14:57 |
| creiht | peluse_: yeah I don't get in until late sunday as well | 14:57 |
| *** zhiyan is now known as zhiyan_ | 15:04 | |
| acoles | gholt: i have been playing with this to show me what changed between versions on a patchset http://paste.openstack.org/show/82196/ | 15:09 |
| *** haomaiw__ has quit IRC | 15:10 | |
| *** kenhui1 has joined #openstack-swift | 15:10 | |
| *** haomaiwang has joined #openstack-swift | 15:10 | |
| * acoles waits for someone to point out a more elegant solution | 15:11 | |
| dfg | on the plus side the review process is so complicated that the actual code change seems pretty straight forward | 15:12 |
| peluse_ | I hear that! | 15:12 |
| creiht | haha | 15:13 |
| *** kenhui has quit IRC | 15:13 | |
| *** haomaiw__ has joined #openstack-swift | 15:14 | |
| acoles | :) | 15:15 |
| * peluse_ is stepping away for 90 min or so... | 15:15 | |
| *** haomaiwang has quit IRC | 15:17 | |
| *** ozialien has joined #openstack-swift | 15:48 | |
| openstackgerrit | paul luse proposed a change to openstack/swift: Make quarantine and tmp dirs in DiskFile Storage Policy Aware https://review.openstack.org/96313 | 15:49 |
| *** Midnightmyth has quit IRC | 15:55 | |
| *** zhiyan_ is now known as zhiyan | 15:56 | |
| *** ozialien has quit IRC | 16:02 | |
| openstackgerrit | Yuan Zhou proposed a change to openstack/swift: Update swift-object-info/swift-get-nodes to be storage policy aware https://review.openstack.org/82734 | 16:04 |
| *** zhiyan is now known as zhiyan_ | 16:05 | |
| *** haomaiw__ has quit IRC | 16:07 | |
| *** haomaiwa_ has joined #openstack-swift | 16:07 | |
| *** mwstorer has joined #openstack-swift | 16:10 | |
| *** Midnightmyth has joined #openstack-swift | 16:13 | |
| *** haomaiw__ has joined #openstack-swift | 16:17 | |
| *** haomaiwa_ has quit IRC | 16:18 | |
| *** shakamunyi has quit IRC | 16:20 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/python-swiftclient: Updated from global requirements https://review.openstack.org/89250 | 16:33 |
| *** Ju has quit IRC | 16:52 | |
| *** shri has joined #openstack-swift | 16:53 | |
| *** NM has quit IRC | 16:55 | |
| *** zhiyan_ is now known as zhiyan | 16:56 | |
| *** as_ has joined #openstack-swift | 17:00 | |
| creiht | peluse_, clayg: Why did we need a separate policy stat table in the container? | 17:00 |
| *** kashyapk has joined #openstack-swift | 17:00 | |
| notmyname | good morning. just got done talking about swift in an HP podcast :-) | 17:04 |
| *** ozialien has joined #openstack-swift | 17:04 | |
| *** nacim has quit IRC | 17:06 | |
| *** zhiyan is now known as zhiyan_ | 17:06 | |
| as_ | does anybody know swiftbrowser? my problem is: when I try to upload something from the swiftbrowser, gets me to that url: https://MYIPHERE:8076/v1/AUTH_system/Test/ | 17:07 |
| as_ | And I get 401 error | 17:07 |
| as_ | Unauthorized | 17:07 |
| as_ | This server could not verify that you are authorized to access the document you requested. | 17:07 |
| as_ | I've tried with CloudBerry, works perfectly.. so this should be swiftbrowser problem | 17:08 |
| notmyname | as_: hmmm | 17:08 |
| notmyname | as_: can you give me a link to the swiftbrowser you are using? (just want to make sure what I'm assuming you are using is actually what you are using) | 17:08 |
| as_ | https://github.com/cschwede/django-swiftbrowser/tree/master/swiftbrowser | 17:09 |
| notmyname | as_: ah yes. you should talk to cschwede_ then :-) | 17:09 |
| as_ | uhm, ok thanks :p | 17:09 |
| notmyname | as_: that is, he's in here. and he's on central euro time (IIRC), so it's kinda late for him now, but he should see it. | 17:10 |
| as_ | will send him an e-mail :D thank you for your time; have a good day! | 17:11 |
| *** as_ has quit IRC | 17:11 | |
| *** miqui has quit IRC | 17:18 | |
| *** erlon has quit IRC | 17:19 | |
| peluse_ | creiht: because we're keeping a row of stats per policy | 17:19 |
| *** NM has joined #openstack-swift | 17:20 | |
| notmyname | creiht: did you see this? how LEDs are made https://learn.sparkfun.com/tutorials/how-leds-are-made/all | 17:21 |
| peluse_ | creiht: also note that the table is in the acct DB (you mentioned container) and is analagous to account_stat | 17:25 |
| *** miqui has joined #openstack-swift | 17:29 | |
| creiht | notmyname: yeah | 17:44 |
| creiht | peluse_: ok, but a single container can only have one policy right? | 17:44 |
| creiht | peluse_: I'm looking at: https://review.openstack.org/#/c/96029/6/swift/container/backend.py | 17:44 |
| creiht | POLICY_STAT_TABLE_CREATE | 17:45 |
| *** nshaikh has quit IRC | 17:52 | |
| * peluse_ is looking | 17:52 | |
| *** NM has quit IRC | 17:55 | |
| peluse_ | creiht: so that's not where that code is supposed to be.... | 17:55 |
| peluse_ | I'm really confused | 17:55 |
| creiht | heh | 17:55 |
| peluse_ | that's not where it is on feature/ec (and not where it I added it).... | 17:56 |
| creiht | odd | 17:56 |
| creiht | I'll leave a note for now and let yall sort it out | 17:56 |
| peluse_ | creiht: thanks, I'm sure clayg knows what's up there | 17:57 |
| *** zhiyan_ is now known as zhiyan | 17:57 | |
| creiht | peluse_: in my full pull of code, the account backend has a create_policy_stat_table | 17:59 |
| peluse_ | creiht: yeah, that's the one I did. | 17:59 |
| peluse_ | creiht: clayg must have added this one in the last few weeks | 18:00 |
| creiht | k | 18:00 |
| creiht | I'll wait for some clarity from clayg | 18:01 |
| *** zaitcev has joined #openstack-swift | 18:01 | |
| *** ChanServ sets mode: +v zaitcev | 18:01 | |
| peluse_ | creiht: yeah, here's the one I did https://review.openstack.org/#/c/96041/ so its totally different | 18:03 |
| peluse_ | and that one ^ is several patches downstream in the set... | 18:04 |
| *** gyee has joined #openstack-swift | 18:04 | |
| *** kashyapk has quit IRC | 18:06 | |
| *** zhiyan is now known as zhiyan_ | 18:06 | |
| *** kashyapk has joined #openstack-swift | 18:07 | |
| *** NM has joined #openstack-swift | 18:10 | |
| *** kashyapk has quit IRC | 18:11 | |
| creiht | notmyname: on a weird side note, I tried recreating all of my notification settings in gerrit. I get emails for updates for every one, except swift proper | 18:20 |
| notmyname | creiht: oh fun. just what you wanted :-) | 18:21 |
| creiht | heh | 18:22 |
| creiht | notmyname: well be clear, the other updates are the other ones that I signed up for | 18:29 |
| creiht | I just don't get the updates for swift which I also tried to sign up for again | 18:29 |
| creiht | I think everyone might have discussed this earlier | 18:40 |
| creiht | but why do we have constants for the policy header keys, but we don't for any of the others? | 18:40 |
| clayg | hrm... was having some computer trouble this morning | 18:47 |
| clayg | might be working now, it seems i've got some comments to go through - thanks all | 18:47 |
| creiht | clayg: howdy :) | 18:48 |
| clayg | creiht: the policy_stat table was added to the container to deal with object rows from different storage policies | 18:51 |
| creiht | hrm | 18:51 |
| creiht | seems like a lot of overhead just for that plus the backwards incompatibility | 18:52 |
| clayg | creiht: yup | 18:53 |
| creiht | because that's just for when you get split brained containers right? | 18:53 |
| clayg | creiht: well... yeah | 18:53 |
| creiht | ok I need to think about that for a bit, it would prefer if we didn't need the extra policy_stat table in the container | 18:54 |
| creiht | s/it/I | 18:54 |
| clayg | creiht: i would have preferred that as well - and fought against it for a long time before I finally gave into the inevitble conclusion | 18:55 |
| creiht | heh | 18:55 |
| creiht | yeah that's why I need to think about it a bit to see if I can come up with a better solution to that issue | 18:55 |
| clayg | creiht: ok, well I'll document the other solutions i explored - maybe you'll have better ideas on how they could have been made to work. | 18:56 |
| creiht | k | 18:56 |
| *** haomaiw__ has quit IRC | 18:58 | |
| *** zhiyan_ is now known as zhiyan | 18:58 | |
| creiht | clayg: what about the constants for the policy headers? | 19:01 |
| creiht | I think this might have been covered earlier | 19:02 |
| creiht | but why do we have them when no other headers have constants? | 19:02 |
| clayg | oh heh, so you wanted to drop the constants and go with string literals - funny. | 19:02 |
| creiht | well, I prefer to be consistent :) | 19:02 |
| creiht | I don't have a real strong preference one way or another | 19:03 |
| creiht | each way has advantages/disadvantages | 19:03 |
| clayg | everyone else seemed to think it was better to have something flake8 could check - notmyname's only concern was that the constants' names didn't indicate they were headers. | 19:03 |
| clayg | I believe the concensus was to keep them, but rename to POLICY_HEADER and POLICY_INDEX_HEADER | 19:04 |
| creiht | heh | 19:04 |
| clayg | I originally was in the "just use string literals" camp and didn't like the names - but i got used to POLICY and POLICY_INDEX and came to not care. | 19:04 |
| creiht | well I would hope that we have unit tests that can test if we typo a header somewhere :) | 19:05 |
| creiht | I feel relying on flake8 for correctness is not helpful | 19:05 |
| gholt | I was in the not giving a flip camp and still am, but unfortunately there seem to be lots of flip givers that each give different flips. ;) | 19:05 |
| creiht | heh | 19:06 |
| creiht | I just really want consistency | 19:06 |
| *** NM has quit IRC | 19:06 | |
| gholt | That's the latest flip given. :D | 19:06 |
| creiht | heh | 19:07 |
| *** zhiyan is now known as zhiyan_ | 19:07 | |
| gholt | I'm imaging clayg changing all the header string literals across all the code to constants. Heheh. | 19:08 |
| clayg | gholt: not gunna happen | 19:08 |
| creiht | hah | 19:08 |
| gholt | Well, the other thing my imagination had was you beating someone Clockwork Orange style, so... | 19:09 |
| creiht | if nobody else cares.... | 19:09 |
| creiht | it will just bug me every time that I look at it :) | 19:09 |
| gholt | clayg: I don't know if you got to see all the confusion around reviewing the various patchsets each with their own dependency tree. But I think I'm good now. I'm just going to work from the last patch here on and then translate where to plop comments I have based on git blame git show etc. | 19:11 |
| clayg | creiht: you put up a patch that changes them and see if folks like it in review. | 19:11 |
| clayg | gholt: yeah gerrits' handling of the "Base" patchset isn't as great as I would like. | 19:11 |
| gholt | I just hope it's not too bad on your end, keeping up with everyone's comments and then figuring out which commit to alter or whatever it is you have do. O.o | 19:13 |
| clayg | gholt: I feel like if I went through each patch change by change and did a rebase on it's parent, review, apply the days changes, review, ... then the Base thing could show you what changed from patch a to c with the intermediate change b as the base which is just there to eat the parent patchset changes | 19:13 |
| clayg | gholt: i'm ok so far, but most of the changes have been fairly astetic - i'm glad we got types out and deprecation in though. | 19:14 |
| gholt | Yeah, definitely do not make your job even harder, esp just on my account. Working from the last patchset's diffs will do quite nicely for me. | 19:14 |
| * creiht takes a break for a bit | 19:23 | |
| *** kutija has quit IRC | 19:29 | |
| zaitcev | "Setting up Storage Policies was very simple, and using them is even simpler." -- From the Billy Mays school of technical writing. | 19:31 |
| zaitcev | I'm Loving It | 19:32 |
| peluse_ | nice! | 19:44 |
| peluse_ | man, I never thought the header name contants would garner so much attention :) creiht, FYI, I introduced those early on because we couldn't decide on a name for the longest time so instead of contantly changing literals everyone I just made those constants. Personally, I'm cool w/however they end up... | 19:46 |
| creiht | hehe | 19:46 |
| creiht | oh I can understand :) | 19:46 |
| peluse_ | s/everyone/everywhere/ | 19:46 |
| notmyname | Alex_Gaynor: around? | 19:47 |
| peluse_ | hey, that reminds me, totally off topic. What's the deal with pypy? I saw some stuff from Alex last year about it, is anyone pursuing a strategy there? Someone at work asked me about it... | 19:48 |
| notmyname | creiht: clayg: I'm fine with them being strings or constants. seemed the consensus yesterday was to keep them that way. mostly I figured it was simpler to keep them and get SP merged rather than bikeshed on it | 19:48 |
| creiht | notmyname: yeah that's fine... the only thing I was arguing for was consistency :) | 19:48 |
| notmyname | peluse_: I think the "strategy" is just to make sure it works under pypy then deployers can use what they want | 19:48 |
| notmyname | creiht: ya. I'm quite sympathetic with you there | 19:49 |
| *** madhuri has quit IRC | 19:50 | |
| peluse_ | notmyname: ah, OK. Does swift run under pypy today? | 19:50 |
| notmyname | peluse_: AFAIK | 19:51 |
| notmyname | peluse_: let me rephrase: there's a strong chance that definitely maybe it does | 19:51 |
| Alex_Gaynor | notmyname: yar | 19:51 |
| peluse_ | clayg: hey man, need me to take care of anything (or anyone)? | 19:51 |
| creiht | peluse_: it is *just* python, so it should *just* work, right :) | 19:52 |
| notmyname | Alex_Gaynor: I'm looking at pypi download stats, and google tells me you wrote a blog about it. do you know if there is any way to get historical info beyond what's on the project page? | 19:52 |
| clayg | peluse_: idk, i'm working on a "the evolution of tracking misplaced objects" for chuck | 19:52 |
| Alex_Gaynor | notmyname: the way you get that info is you ping dstufft in #openstack-infra or #pypa and ask nicely | 19:52 |
| peluse_ | clayg: OK, well just hollar if you need me to take care of anything | 19:53 |
| clayg | peluse_: I think there's doc comments if you wanna get me a diff - really anything that will apply on b4d4ba75d31 would be great if you wannt just check out the end of the patch change and hack on another commit | 19:53 |
| notmyname | Alex_Gaynor: heh, ok thanks :-) | 19:54 |
| tdasilva | hey guys...just fyi..while reviewing and testing SP, I've been able to setup swiftonfile as a storage policy, pretty cool | 19:56 |
| notmyname | tdasilva: nice! | 19:56 |
| peluse_ | clayg: which is b4d4ba75d31? To be clear, want me to jsut address comments on 96026 (docs) and send a diff or something different? | 19:57 |
| tdasilva | if anyone is interested, this pull request https://github.com/swiftonfile/swiftonfile/pull/19 is the work that we are doing to make swiftonfile a SP | 19:57 |
| *** zhiyan_ is now known as zhiyan | 19:59 | |
| clayg | peluse_: check out the last patch in the series and make changes addressing comments in the reviews. Get me that diff. THANKS! | 19:59 |
| clayg | peluse_: (i think the last one is something something Reaper - which seems fitting) | 20:00 |
| peluse_ | clayg: got it, cool | 20:01 |
| *** Thurloat has quit IRC | 20:04 | |
| clayg | creiht: https://gist.github.com/clayg/84e11b94735016c1c341 | 20:07 |
| clayg | gholt: ^ you should probably give it a read too | 20:07 |
| * clayg knew this day would come | 20:07 | |
| clayg | i'm so sorry guys | 20:07 |
| gholt | Ohshitwalloftext thiscan'tbegood | 20:08 |
| *** zhiyan is now known as zhiyan_ | 20:08 | |
| creiht | holy cow | 20:08 |
| creiht | clayg: hehe thanks for the context | 20:08 |
| gholt | Heheh, oh yeah, I remember asking for this. :) | 20:08 |
| *** spredzy has joined #openstack-swift | 20:11 | |
| *** otoolee has quit IRC | 20:15 | |
| *** jeblair_ has joined #openstack-swift | 20:20 | |
| *** jeblair has quit IRC | 20:21 | |
| *** jeblair_ is now known as jeblair | 20:22 | |
| *** tdasilva has left #openstack-swift | 20:22 | |
| *** otoolee has joined #openstack-swift | 20:23 | |
| *** kenhui1 has quit IRC | 20:23 | |
| zaitcev | I only started to understand what Clay meant in that gist in the penultimate paragraph. | 20:25 |
| * clayg goes to look up penultimate | 20:25 | |
| notmyname | next-to-last | 20:25 |
| clayg | too late - google is faster | 20:26 |
| zaitcev | I was kinda of wondering. Russian and Japanese have a word for "the day after tomorrow", "poslezavtra" and "asatte" respectively. Why don't English? Or does it? | 20:27 |
| clayg | overmorrow? | 20:28 |
| zaitcev | something like that | 20:28 |
| zaitcev | http://english.stackexchange.com/questions/45412/how-obsolete-is-the-word-overmorrow | 20:29 |
| *** ozialien has quit IRC | 20:30 | |
| *** NM has joined #openstack-swift | 20:34 | |
| peluse_ | clayg: just emailed you resolution for latest comments on doc patch (based off of the last in the chain, the acct reaper) | 20:36 |
| peluse_ | zaitcev: how about "Sunday" :) | 20:37 |
| peluse_ | clayg: FYI I also fixed a few other grammar/spelling things that were somehow still lingering... | 20:38 |
| *** miqui has quit IRC | 20:38 | |
| clayg | peluse_: i think i keep adding more everytime I try to write something someone asked me to elaborate on | 20:39 |
| peluse_ | clayg: yeah, I managed to add a few when I added the section on defecation | 20:39 |
| peluse_ | argh, deprecation :) | 20:39 |
| *** nthacker has joined #openstack-swift | 20:39 | |
| clayg | oh my | 20:39 |
| spredzy | I am hitting an issue for every swift command I try to run : IndexError: list index out of range | 20:39 |
| spredzy | I am not sure what it is due to. I create d the ring fine, added the zone and rebalanced it just fine | 20:40 |
| nthacker | Hi All, How do i run only a single unit test ? | 20:40 |
| spredzy | but when I try to run a swift stat | 20:40 |
| spredzy | it crashed an error 500 with IndexError: list index out of range | 20:40 |
| peluse_ | single unit test example: nosetests --nocapture --nologcapture test/unit/proxy/controllers/test_obj.py:TestObjController.test_COPY_simple | 20:41 |
| spredzy | swift-ring-build {account,container,object}.builder shows correct output | 20:41 |
| zaitcev | nthacker: (cd test/unit && nosetests common/test_wsgi.py:TestWSGI.test_run_server_conf_dir) | 20:42 |
| nthacker | zaitcev: thanks! | 20:42 |
| clayg | peluse_: git diff > mychanges.patch works well with git apply mychanges.patch | 20:43 |
| peluse_ | clayg: cool, so what I sent you worked OK then? | 20:44 |
| clayg | peluse_: yeah i'll make it work - thanks! | 20:44 |
| peluse_ | :) | 20:44 |
| zaitcev | spredzy: The only time I hit that was when I had an unclean /srv/node/dev with extraneous index directories and I used a ring with a smaller size than apparenty it had before. | 20:45 |
| zaitcev | spredzy: You probably have to figure your specific cause out for yourself, sorry. | 20:46 |
| spredzy | zaitcev, hehe ok. will keep digging then thanks for the answer | 20:46 |
| clayg | peluse_: sweet those applied great | 20:46 |
| spredzy | quick q. should the device be a minimal size ? | 20:46 |
| spredzy | ue. 1G, 100M ? | 20:47 |
| spredzy | for testing purpose only | 20:47 |
| peluse_ | clayg: rock n roll | 20:47 |
| peluse_ | spredzy: have you already tried setting up an all in one to get familiar with everything? http://docs.openstack.org/developer/swift/development_saio.html | 20:55 |
| *** zhiyan_ is now known as zhiyan | 20:59 | |
| *** ashish_ has joined #openstack-swift | 21:01 | |
| *** openstackgerrit has quit IRC | 21:01 | |
| *** ashish_ has quit IRC | 21:05 | |
| *** zhiyan is now known as zhiyan_ | 21:09 | |
| *** swat30 has quit IRC | 21:18 | |
| *** kenhui has joined #openstack-swift | 21:28 | |
| *** kashyapk has joined #openstack-swift | 21:44 | |
| *** kashyapk has quit IRC | 21:49 | |
| *** kenhui has quit IRC | 21:56 | |
| *** zhiyan_ is now known as zhiyan | 22:00 | |
| *** judd7 has quit IRC | 22:02 | |
| *** zhiyan is now known as zhiyan_ | 22:09 | |
| *** Midnightmyth has quit IRC | 22:11 | |
| *** mkollaro has quit IRC | 22:13 | |
| clayg | oh wow, it's quiet | 22:16 |
| peluse_ | yeah, Fri afternoon... | 22:16 |
| peluse_ | beer-thirty here soon | 22:16 |
| notmyname | clayg: peluse_: well my meetings with the sales and marketing people are done now, so maybe I can start on the technical part of my day :-) | 22:23 |
| gholt | clayg: For some reason I completely missed that that query is based on ROWID and has a limited start and count. So, just, uhm, pretend I never said anything there. | 22:25 |
| gholt | Probably shouldn't review code when feeling ill. Never really works out well for anyone. ;) | 22:28 |
| *** gyee has quit IRC | 22:28 | |
| *** openstackgerrit has joined #openstack-swift | 22:31 | |
| *** openstackgerrit has quit IRC | 22:35 | |
| *** openstackgerrit has joined #openstack-swift | 22:35 | |
| clayg | gholt: it's fine man, thanks for looking at that | 22:42 |
| clayg | creiht: you wanna talk about *args? | 22:42 |
| clayg | ... yeah me either | 22:43 |
| peluse_ | clayg: what is noga? | 22:43 |
| gholt | Hehe! | 22:43 |
| *** NM has quit IRC | 22:43 | |
| notmyname | peluse_: noqa ? | 22:44 |
| gholt | I feel a little bad though because, looking at all that code more, it really looks like you spent a lot of time and thought making sure long locking queries would never happen. But ah well, I'm smarter now for having said something stupid. :) | 22:44 |
| clayg | gholt: no i was just scared to death of breaking something - don't get all "oh i'm sure clay did this right" on me now | 22:45 |
| peluse_ | notmyname: ask clayg? I dunno :) In a few comments in the reconciler | 22:45 |
| gholt | Well, you know that won't happen anyway. We don't even assume that about ourselves. | 22:46 |
| clayg | peluse_: oh it probably is a typo then - yeah I was probably going for noqa | 22:46 |
| peluse_ | which is? | 22:46 |
| notmyname | peluse_: at the end of the patch chain I can't find "noga" anywhere | 22:46 |
| clayg | around some bare excepts maybe? I'm always thinking that thing we got meged into hacking will get release some day. | 22:46 |
| peluse_ | yeah, its noqa | 22:46 |
| clayg | peluse_: it means no QA - skip flake8 checks | 22:46 |
| peluse_ | ahh | 22:47 |
| notmyname | peluse_: IOW "I know this will fail a flake8 test and that's ok" | 22:47 |
| gholt | Yeah, and there are some times when coverage will insist you haven't covered a line even though you've verified you truly have. | 22:47 |
| gholt | Usually, pass is the culprit | 22:47 |
| clayg | notmyname: I don't think we even run the flake8 check i'm noqa'ing against - but you know - forward thinking! it's like py3 support. | 22:48 |
| peluse_ | did not know that... thanks. | 22:48 |
| *** jergerber has quit IRC | 22:52 | |
| clayg | peluse_: I need add doc sections for the container-reconciler config options | 22:52 |
| clayg | notmyname: did creiht make any progress arguing for string literals and consistency? | 22:53 |
| clayg | acoles: thanks for the reminder on the X-Backend-Timestamp thing on object 404 | 22:53 |
| peluse_ | clayg: just left you a reminder on the review to do that :) | 22:54 |
| clayg | thanks | 22:54 |
| *** zhiyan_ is now known as zhiyan | 23:01 | |
| *** zhiyan is now known as zhiyan_ | 23:10 | |
| *** nsquare has joined #openstack-swift | 23:13 | |
| *** diegows has quit IRC | 23:13 | |
| clayg | peluse_: oh right i need to review the tmp/q change again | 23:19 |
| peluse_ | ja | 23:30 |
| clayg | comin' at ya | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Update container storage_policy_index when auditing https://review.openstack.org/96352 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add functional tests for Storage Policy https://review.openstack.org/96042 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Fixes probe tests with non-zero default storage policy https://review.openstack.org/96353 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Update swift-container-info to be storage policy aware https://review.openstack.org/96043 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Merge container storage_policy_index https://review.openstack.org/96040 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Support to Accounts https://review.openstack.org/96041 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Support to list_endpoints https://review.openstack.org/96046 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Support to Container Sync https://review.openstack.org/96047 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Enqueue misplaced objects during container replication https://review.openstack.org/96044 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add LRUCache to common.utils https://review.openstack.org/96045 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Support to ssync https://review.openstack.org/96034 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Put X-Timestamp in object 404 responses https://review.openstack.org/96035 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Support to the Auditor https://review.openstack.org/96032 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add storage policy support for the Replicator https://review.openstack.org/96033 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add container-reconciler daemon https://review.openstack.org/96038 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add reconciler probetest outline https://review.openstack.org/96039 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Extend interface on InternalClient https://review.openstack.org/96036 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Extend direct_client https://review.openstack.org/96037 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Support https://review.openstack.org/96027 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Documentation https://review.openstack.org/96026 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy support to Containers https://review.openstack.org/96029 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Update FakeRing and FakeLogger https://review.openstack.org/96028 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy support to Object Updates https://review.openstack.org/96031 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy support to Object Server https://review.openstack.org/96030 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy support to the Account Reaper https://review.openstack.org/96049 | 23:31 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Support to Recon Middleware https://review.openstack.org/96048 | 23:31 |
| clayg | everyone have a good weekend - see you next week | 23:32 |
| peluse_ | awesome clayg, see ya Mon | 23:32 |
| clayg | oh, and sorry for all the updates gholt - maybe you can do a working session on how to review these changes or we can come up with something different than maintaing all of these... | 23:32 |
| *** NM has joined #openstack-swift | 23:33 | |
| clayg | I did notice that if you go to the end and select the last base you can see all of the last bits changes (if you just wanna see if your comment was addressed) | 23:33 |
| *** ozialien has joined #openstack-swift | 23:38 | |
| gholt | clayg: Yeah, working from the end seems best. | 23:40 |
| *** NM has quit IRC | 23:41 | |
| *** NM has joined #openstack-swift | 23:41 | |
| *** NM has quit IRC | 23:51 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!