| *** itlinux has quit IRC | 00:03 | |
| *** chlong has joined #openstack-swift | 00:41 | |
| *** thurloat is now known as thurloat_isgone | 00:46 | |
| *** baojg has joined #openstack-swift | 01:02 | |
| *** garthb has joined #openstack-swift | 01:34 | |
| *** tsimp06 has quit IRC | 01:34 | |
| *** hezhiqiang has joined #openstack-swift | 01:38 | |
| *** haomaiwa_ has joined #openstack-swift | 02:07 | |
| *** itlinux has joined #openstack-swift | 02:14 | |
| *** DericHorn-HP has joined #openstack-swift | 02:18 | |
| *** haomaiwa_ has quit IRC | 02:18 | |
| *** DericHorn-HP has quit IRC | 02:20 | |
| *** hezhiqia_ has joined #openstack-swift | 02:26 | |
| *** esker has quit IRC | 02:26 | |
| *** hezhiqia_ has quit IRC | 02:30 | |
| *** hezhiqiang has quit IRC | 02:30 | |
| *** DericHorn-HP has joined #openstack-swift | 02:32 | |
| *** mragupat has joined #openstack-swift | 02:34 | |
| *** mragupat_ has joined #openstack-swift | 02:35 | |
| *** DericHorn-HP has quit IRC | 02:36 | |
| notmyname | good evening | 02:36 |
|---|---|---|
| *** mragupat has quit IRC | 02:39 | |
| *** haomaiwa_ has joined #openstack-swift | 02:48 | |
| *** baojg has quit IRC | 02:52 | |
| *** sanchitmalhotra has joined #openstack-swift | 02:53 | |
| *** mragupat_ has quit IRC | 02:56 | |
| *** silor has joined #openstack-swift | 02:56 | |
| *** baojg has joined #openstack-swift | 02:59 | |
| *** baojg has quit IRC | 02:59 | |
| *** haomaiwa_ has quit IRC | 03:01 | |
| *** haomaiwang has joined #openstack-swift | 03:01 | |
| *** silor has joined #openstack-swift | 03:05 | |
| *** sanchitmalhotra1 has joined #openstack-swift | 03:20 | |
| mattoliverau | notmyname: o/ | 03:20 |
| notmyname | hi | 03:20 |
| mattoliverau | Was lunching when you came online :) | 03:20 |
| notmyname | I'm trying to figure out what's going on after being out a week | 03:20 |
| mattoliverau | Back at home with the fam this weekend? | 03:20 |
| notmyname | yeah, we went to the beach today | 03:20 |
| jrichli | mattoliverau, notmyname: hello | 03:21 |
| notmyname | https://twitter.com/notmyname/status/645661945283002368 | 03:21 |
| jrichli | beach! I am so jealous | 03:21 |
| mattoliverau | \o/ beach! | 03:21 |
| mattoliverau | jrichli: o/ | 03:21 |
| jrichli | I am gonna work from Dallas next week ... I get to see some friends and go to some fav places to eat | 03:22 |
| jrichli | this week | 03:22 |
| *** sanchitmalhotra has quit IRC | 03:22 | |
| mattoliverau | Cool! | 03:22 |
| notmyname | cool. I hope yiou have a good time | 03:22 |
| jrichli | notmyname: it may have been just a glitch this weekend, but I uploaded a couple patches to feature/crypto, and there was no IRC notification. | 03:23 |
| notmyname | jrichli: if gerrit was offline or the bot was offline, no notice would have been given | 03:25 |
| notmyname | happens from time to time | 03:25 |
| notmyname | as long as you see it in gerrit, everything is ok (from our perspective) | 03:25 |
| *** baojg has joined #openstack-swift | 03:26 | |
| jrichli | ok, I figured it might be something like that. Just thought I'd mention it. | 03:26 |
| notmyname | mattoliverau: what are you working on these days? and how is that work going? | 03:27 |
| * notmyname is not a fan of specs these days. we should talk about it in tokyo | 03:28 | |
| *** itlinux has quit IRC | 03:29 | |
| notmyname | jrichli: we should do the merge master to encryption branch thing | 03:29 |
| notmyname | jrichli: have you done that before? | 03:29 |
| jrichli | acoles usually does that. the last one was a couple weeks ago. | 03:30 |
| notmyname | I was hoping to spread the knowledge around ;-) | 03:31 |
| jrichli | oh, ok. Do I need any special permissions? | 03:31 |
| jrichli | it seems like I recall a discussion about this, and I thought it was said that you had to be core | 03:32 |
| jrichli | but maybe that was something else about what was being done | 03:32 |
| notmyname | I don't remember (and it seems the docs I had saved are now a stale link) | 03:34 |
| notmyname | but yeah, the basic idea is to submit a merge commit for review instead of a "normal" commit | 03:34 |
| jrichli | ok. I can give it a shot :-) Not tonight, though. I am just doing a little Japanese and then going to bed. Then tomorrow, I wont be working until night time. | 03:36 |
| openstackgerrit | John Dickinson proposed openstack/swift: Merge branch 'master' into feature/crypto https://review.openstack.org/225540 | 03:37 |
| notmyname | well, there :-) | 03:37 |
| notmyname | just makign sure I remembered how. | 03:37 |
| jrichli | wow, that was fast. i guess there were no complicated merges to resolve? | 03:37 |
| notmyname | none | 03:37 |
| notmyname | this time ;-) | 03:37 |
| jrichli | great. thx! | 03:38 |
| notmyname | I can abandon and let you do it to learn | 03:38 |
| notmyname | your call | 03:38 |
| jrichli | abandon. I would like to try. good time, since there are no conflicts :-) | 03:39 |
| notmyname | done | 03:39 |
| notmyname | want me to walk you through it? | 03:39 |
| *** hezhiqiang has joined #openstack-swift | 03:40 | |
| jrichli | ok. I guess Rosetta Stone can stay paused ;-) | 03:40 |
| notmyname | heh | 03:40 |
| notmyname | jrichli: step one is to make sure master and feature/crypto (or origin/feature/crypto) are up to date | 03:40 |
| notmyname | basicall `git fetch --all` and then eg `git merge origin master` as necessary | 03:41 |
| notmyname | (let me know when you're ready for the next step) | 03:42 |
| *** garthb has quit IRC | 03:44 | |
| *** sanchitmalhotra has joined #openstack-swift | 03:45 | |
| *** hezhiqiang has quit IRC | 03:45 | |
| *** sanchitmalhotra1 has quit IRC | 03:47 | |
| jrichli | if I do the merge command you have above while on crypto branch, then it will be the direction I want? | 03:47 |
| notmyname | yeah, but don't do that yet ;-) | 03:48 |
| notmyname | or there's a safer way than to do it directly | 03:49 |
| jrichli | 52 files changed, 2450 insertions(+), 395 deletions(-) | 03:49 |
| jrichli | already done: 52 files changed, 2450 insertions(+), 395 deletions(-) | 03:49 |
| jrichli | is that what you had? | 03:49 |
| notmyname | yeah. exactly the same | 03:49 |
| jrichli | what's the safer way? | 03:49 |
| notmyname | so I had done it on a separate branch off of feature/crypto | 03:50 |
| jrichli | ah, right. i usually would have thought of that | 03:50 |
| notmyname | ok, so now the last step | 03:50 |
| notmyname | well, 2 steps | 03:50 |
| notmyname | you need to get a change-id on the commit | 03:50 |
| jrichli | it ok i didn't make a branch for the review? | 03:50 |
| notmyname | yeah | 03:51 |
| notmyname | so `git commit --amend) | 03:51 |
| notmyname | so `git commit --amend` | 03:51 |
| notmyname | then save what's give with no changes (if you're ok with the commit message) | 03:51 |
| *** hezhiqiang has joined #openstack-swift | 03:51 | |
| notmyname | that will run the hook to add the commit id | 03:52 |
| jrichli | sure. i guess it is like rebasing, mostly. except for the merge command :-) | 03:52 |
| notmyname | yeah | 03:52 |
| mattoliverau | notmyname: sharding Poc take 3 or is it 4. Debugging in one of my larger test clusters now.. Hope to start some initial benchmarking this week. (Also need to update spec, I know you'd love that) | 03:52 |
| notmyname | ok, so the last part is the one that has to be done right. otherwise you make a lot of work for yourself, potentially | 03:53 |
| jrichli | k | 03:53 |
| notmyname | jrichli: `git review -R` | 03:53 |
| jrichli | what does the -R mean? | 03:53 |
| notmyname | the -R tells git-review to not do a rebase of the commit on what's already there | 03:53 |
| notmyname | ie you get the merge commit itself instead of all the commits that you just merged in individually | 03:53 |
| jrichli | oic | 03:54 |
| notmyname | mattoliverau: great :-) | 03:54 |
| mattoliverau | Other then that just getting over being sick.. Again.. Now that brain is working i need to give reviews some love. But once I'm at the benchmarking stage I'll have a lot more time to review and bug fix :) | 03:54 |
| notmyname | mattoliverau: over the next couple of weeks, anything you can do to help move patches along for the liberty release would be appreciated. EC patches especially | 03:54 |
| jrichli | I got : (you are not allowed to upload merges) | 03:55 |
| notmyname | jrichli: ah | 03:55 |
| notmyname | ok then | 03:55 |
| notmyname | :-( | 03:55 |
| notmyname | well, now we know | 03:55 |
| jrichli | but its good to know how its done | 03:55 |
| jrichli | thx! | 03:55 |
| mattoliverau | notmyname: will do | 03:55 |
| notmyname | jrichli: thanks for trying | 03:56 |
| jrichli | sure! thanks for walking me through. BTW, can we see the t-shirts yet? | 03:56 |
| notmyname | no ;-) | 03:57 |
| jrichli | darn. do we have to wait until Tokyo? | 03:57 |
| notmyname | I'll order them in the morning (or that's the plan). we'll see how they turn out | 03:58 |
| jrichli | ok. I am sure they will be great! | 03:59 |
| *** haomaiwang has quit IRC | 04:01 | |
| *** haomaiwang has joined #openstack-swift | 04:01 | |
| *** hezhiqiang has quit IRC | 04:05 | |
| *** hezhiqiang has joined #openstack-swift | 04:08 | |
| notmyname | torgomatic: I love the topic message on https://review.openstack.org/#/c/212824/ | 04:08 |
| *** wer_ has quit IRC | 04:19 | |
| *** hezhiqiang has quit IRC | 04:22 | |
| *** sanchitmalhotra1 has joined #openstack-swift | 04:25 | |
| *** sanchitmalhotra has quit IRC | 04:27 | |
| *** ppai has joined #openstack-swift | 04:28 | |
| *** ppai has quit IRC | 04:40 | |
| *** DericHorn-HP has joined #openstack-swift | 04:40 | |
| *** sanchitmalhotra has joined #openstack-swift | 04:45 | |
| *** sanchitmalhotra1 has quit IRC | 04:47 | |
| *** DericHorn-HP has quit IRC | 04:48 | |
| *** baojg has quit IRC | 04:50 | |
| *** garthb has joined #openstack-swift | 04:54 | |
| *** sanchitmalhotra1 has joined #openstack-swift | 04:58 | |
| *** nadeem has joined #openstack-swift | 04:59 | |
| *** hezhiqiang has joined #openstack-swift | 04:59 | |
| *** sanchitmalhotra has quit IRC | 04:59 | |
| *** haomaiwang has quit IRC | 05:01 | |
| *** haomaiwang has joined #openstack-swift | 05:01 | |
| *** itlinux has joined #openstack-swift | 05:03 | |
| *** patchbot has quit IRC | 05:06 | |
| *** baojg has joined #openstack-swift | 05:15 | |
| *** sanchitmalhotra has joined #openstack-swift | 05:19 | |
| *** sanchitmalhotra1 has quit IRC | 05:21 | |
| *** nadeem has quit IRC | 05:27 | |
| *** nadeem has joined #openstack-swift | 05:28 | |
| *** patchbot has joined #openstack-swift | 05:28 | |
| notmyname | patchbot has been upgraded | 05:30 |
| notmyname | now you can reference patch 1234 and patch 89573 in the same line and you'll get a link to both | 05:30 |
| patchbot | notmyname: https://review.openstack.org/#/c/1234/ | 05:30 |
| patchbot | notmyname: https://review.openstack.org/#/c/89573/ | 05:30 |
| notmyname | like that ;-) | 05:30 |
| *** garthb has quit IRC | 05:33 | |
| tamizh_geek | notmyname: clayg: hello, do you guys remember a patch trying to fix the error response content types? I have a new patch regarding that. This one could be much bigger, touching things all over the code base. Would like to have some early feedback on that. | 05:33 |
| tamizh_geek | whether I'm doing it right? That, this is needed at all? etc etc.. | 05:34 |
| tamizh_geek | please help! | 05:34 |
| notmyname | tamizh_geek: yeah, I saw that you had submitted a few. thank you | 05:34 |
| tamizh_geek | patch is 225499 | 05:34 |
| notmyname | take out the "is" in that line, and patchbot will make it linky for you ;-) | 05:34 |
| tamizh_geek | https://review.openstack.org/#/c/225499/ | 05:34 |
| *** jrichli has quit IRC | 05:34 | |
| tamizh_geek | ah! thats what I was wondering :-) | 05:35 |
| tamizh_geek | thanks for the hint! | 05:35 |
| notmyname | fresh on my mind since I just added some functionality to that bot | 05:35 |
| notmyname | tamizh_geek: I'm interested in it, and I'd like to look, but let me be honest with what's going on | 05:35 |
| notmyname | tamizh_geek: we've pretty much got 2 weeks to get the Liberty release banged into shape. that will take priority | 05:36 |
| *** itlinux has quit IRC | 05:36 | |
| *** mahatic has joined #openstack-swift | 05:36 | |
| tamizh_geek | notmyname: I understand. But also, I don't want to proceed on this without knowing what I'm doing is right. Probably, I will keep waiting on further development till I receive some feedback from you guys. | 05:38 |
| notmyname | tamizh_geek: ack. from a quick glance, getting the pep8 would be pretty simple to go ahead and work on | 05:41 |
| *** hezhiqia_ has joined #openstack-swift | 05:41 | |
| notmyname | tamizh_geek: and what you'd need to pay attention to most is the functional tests (and you'd need new ones) | 05:41 |
| notmyname | tamizh_geek: those will show what works and doesn't from a client perspective | 05:41 |
| notmyname | IIRC we've changed responses in the patch to be more in line with the request (eg the Accept header). so I think this kind of change will likely be ok | 05:42 |
| tamizh_geek | notmyname: Thanks! That's the path I'm taking. I'm going through each func test, and passing the JSON header to it and see if it respects it. If not, tweak the code to add necessary changes. In the process, I'm also making sure the old ones still pass. | 05:43 |
| mahatic | good morning | 05:44 |
| tamizh_geek | notmyname: Yeah right! That without breaking backward compatibility. Thanks. This gives some confidence :) | 05:44 |
| notmyname | tamizh_geek: at first glance, it seems like a good idea that you should pursue. however, other cores have been know to (correctly, more often than not) disagree with me ;-) | 05:44 |
| notmyname | mahatic: hello | 05:44 |
| *** hezhiqiang has quit IRC | 05:44 | |
| mahatic | notmyname: hi! how are you? | 05:45 |
| * mahatic was reading scrollback | 05:45 | |
| notmyname | mahatic: pretty good for the most part. it's hot at home today, and I don't have AC, so I'm not looking forward to trying to sleep ;-) | 05:45 |
| tamizh_geek | notmyname: hmmm., clayg had some interest in this patch earlier. I'll probably wait for his feedback as well. | 05:46 |
| mahatic | notmyname: ow. SFO can be that hot? | 05:46 |
| notmyname | mahatic: mostly it's the lack of AC ;-) | 05:46 |
| notmyname | mahatic: but yeah, it was in the upper 80s to low 90s (depending on the weather station you checked) today | 05:47 |
| notmyname | and if there isn't a cross breeze, then the house gets hot. it's 80.41F/26.89C in my office right now | 05:47 |
| mahatic | I see. 26 isn't that bad for us over here ;) but why no AC? all nighter then? ;) | 05:48 |
| notmyname | mahatic: very few houses have AC here. normally it's much cooler year-round. it's been very hot this year | 05:49 |
| notmyname | tamizh_geek: is https://review.openstack.org/#/c/204196/ ready for review? should it be rolled into your other one? | 05:50 |
| mahatic | notmyname: oh okay. yeah, that's right about cross breeze. makes a lot of difference (talking from expereince) | 05:50 |
| notmyname | mahatic: yeah, and I understand eye-rolling I get when I complain about it being 80. I lived most of my life where it is *much* hotter. but I had AC there | 05:51 |
| tamizh_geek | notmyname: I actually made 225499 dependent on 204196. Is that the correct way to roll in related patches ? | 05:52 |
| mahatic | notmyname: well, looks like this year weather should get you used to this and no more complaining ;) | 05:52 |
| tamizh_geek | Just saying.. Chennai/India is at 34C right now ;-) | 05:53 |
| mahatic | notmyname: we're talking about weather, we're doing a british thing :D | 05:54 |
| notmyname | :-) | 05:54 |
| mahatic | tamizh_geek: ah, chennai is super hot! | 05:55 |
| notmyname | I did go to the beach today. water temperature looks like it was about 16C (61F) | 05:55 |
| notmyname | that was really cold! | 05:56 |
| tamizh_geek | mahatic: hehe. this is semi-summer for us. The summer temperature goes up to 42/43C | 05:56 |
| mahatic | I first learned it was a british trait from P.G. wodehouse | 05:56 |
| *** nadeem has quit IRC | 05:57 | |
| mahatic | notmyname: extremes huh. but nice! | 05:58 |
| *** eranrom has joined #openstack-swift | 05:59 | |
| notmyname | tamizh_geek: yeah, you did the dependencies right (the logistics in git/gerrit). but i'm wondering if you need that. does the first patch provide something on its own that is good without the second patch? | 05:59 |
| notmyname | tamizh_geek: if not, then including it in the 2nd patch (and just having one patch instead of a dependency chain) is a better way to go | 06:00 |
| *** haomaiwang has quit IRC | 06:01 | |
| tamizh_geek | notmyname: yes it actually fixes the bug 1182155 - but for the bigger picture it brings in a inconsistency in JSON error responses. Right now, no endpoint respects Accept:application/json. After that patch some will, and some will not. | 06:01 |
| openstack | bug 1182155 in OpenStack Object Storage (swift) "Error responses ignore accept header" [Undecided,In progress] https://launchpad.net/bugs/1182155 - Assigned to Azhagu Selvan SP (tamizhgeek) | 06:01 |
| *** haomaiwang has joined #openstack-swift | 06:01 | |
| tamizh_geek | ok then, I will squash them into a single patch and submit again. | 06:02 |
| tamizh_geek | I was thinking, since its too big let me split them to multiple patches for easy review. | 06:02 |
| notmyname | tamizh_geek: it's up to you | 06:03 |
| notmyname | ok, I'm off to bed. good night | 06:03 |
| tamizh_geek | thanks notmyname, good night :) | 06:03 |
| mahatic | notmyname: good night! | 06:08 |
| *** haomaiwang has quit IRC | 06:12 | |
| *** SkyRocknRoll has joined #openstack-swift | 06:41 | |
| mattoliverau | Someone to remind me to never go down the python unicode investigation.. that's an hour of my life I'm never getting back and still undercided how I feel about a patchset :P | 06:52 |
| mahatic | mattoliverau: lol, which patchset is that? I'm actually debugging something to do with unicode (in encryption). thanks for the encouragement :P | 06:54 |
| mattoliverau | mahatic: lol, sorry :P | 06:55 |
| *** geaaru has joined #openstack-swift | 06:55 | |
| *** sanchitmalhotra1 has joined #openstack-swift | 07:04 | |
| *** SkyRocknRoll has quit IRC | 07:04 | |
| mattoliverau | mahatic: just one about a change in replacing how we parse MIME headers, patch 203304. I started down trying/testing unicode (wrong rabbit hole, and the problem). Ended up reading more of the RFC which says the headers will be in ASCII (great) but can be encoded a different way which can be done with email.header.Header, but email.parser (what the patch is using) doesn't decode it. I don't think we support this | 07:05 |
| mattoliverau | encoding so probably isn't a big deal.. but I guess I should go confirm that the old library doen't also. | 07:05 |
| patchbot | mattoliverau: https://review.openstack.org/#/c/203304/ | 07:05 |
| *** sanchitmalhotra has quit IRC | 07:06 | |
| *** rledisez has joined #openstack-swift | 07:07 | |
| *** SkyRocknRoll has joined #openstack-swift | 07:09 | |
| *** haomaiwang has joined #openstack-swift | 07:09 | |
| *** jamielennox is now known as jamielennox|away | 07:09 | |
| *** trifon has joined #openstack-swift | 07:10 | |
| *** chlong has quit IRC | 07:17 | |
| mahatic | mattoliverau: I see. Looking deeper into that patch, this looks informative | 07:17 |
| *** mac_ified has quit IRC | 07:49 | |
| *** hrou has quit IRC | 07:55 | |
| *** haomaiwang has quit IRC | 08:01 | |
| *** haomaiwang has joined #openstack-swift | 08:01 | |
| *** T0m_ has joined #openstack-swift | 08:12 | |
| *** T0m_ has left #openstack-swift | 08:12 | |
| *** kei_yama has quit IRC | 08:14 | |
| *** kei_yama has joined #openstack-swift | 08:15 | |
| *** janonymous has joined #openstack-swift | 08:16 | |
| *** kei_yama has quit IRC | 08:18 | |
| *** haomaiwang has quit IRC | 08:21 | |
| *** kei_yama has joined #openstack-swift | 08:23 | |
| *** acoles_ is now known as acoles | 08:35 | |
| *** chlong has joined #openstack-swift | 08:43 | |
| *** aix has quit IRC | 08:54 | |
| *** nadeem has joined #openstack-swift | 09:06 | |
| *** kei_yama has quit IRC | 09:08 | |
| *** kei_yama has joined #openstack-swift | 09:09 | |
| openstackgerrit | Alistair Coles proposed openstack/swift: Fix .coveragrc to prevent nose tests error https://review.openstack.org/225591 | 09:12 |
| acoles | cschwede: i think the gate is broken without this fix ^^, please review if you have time. | 09:13 |
| acoles | mattoliverau: kota if you are still around ^^ | 09:13 |
| cschwede | acoles: good morning! alright, going to have a look right now | 09:13 |
| acoles | cschwede: thanks! | 09:13 |
| cschwede | wow, that’s a long patch. will require quite some time ;) | 09:13 |
| acoles | cschwede: saw it here btw: https://review.openstack.org/#/c/225540/ | 09:14 |
| cschwede | ouch | 09:14 |
| acoles | heh. yes its long, that horizontal line has to travel all the way from - to _ :) | 09:15 |
| cschwede | ok, i +2 and +A’ed the patch to unblock the gate - i think this is safe in this case | 09:16 |
| acoles | cschwede: cool. thank you | 09:17 |
| cschwede | acoles: you’re welcome! now we just have to wait for jenkins to pass too | 09:17 |
| *** haomaiwang has joined #openstack-swift | 09:20 | |
| *** haomaiwang has quit IRC | 09:21 | |
| *** haomaiwa_ has joined #openstack-swift | 09:21 | |
| *** kei_yama has quit IRC | 09:22 | |
| *** kei_yama has joined #openstack-swift | 09:22 | |
| acoles | cschwede: jenkins has passed py27 job so at least we didn't break anything ;) | 09:23 |
| *** haomaiwa_ has quit IRC | 09:23 | |
| *** haomaiwa_ has joined #openstack-swift | 09:23 | |
| *** aix has joined #openstack-swift | 09:25 | |
| cschwede | acoles: yep, saw that. according to http://coverage.readthedocs.org/en/latest/config.html#report it also makes more sense to use an underscore than a hyphen | 09:28 |
| *** baojg has quit IRC | 09:35 | |
| *** eranrom has quit IRC | 09:35 | |
| *** nadeem has quit IRC | 09:38 | |
| *** silor has quit IRC | 09:52 | |
| *** mahatic has quit IRC | 09:57 | |
| *** haomaiwa_ has quit IRC | 10:01 | |
| *** haomaiwa_ has joined #openstack-swift | 10:01 | |
| *** kei_yama has quit IRC | 10:03 | |
| *** baojg has joined #openstack-swift | 10:06 | |
| *** baojg has quit IRC | 10:07 | |
| *** hezhiqia_ has quit IRC | 10:20 | |
| *** km has quit IRC | 10:24 | |
| openstackgerrit | Merged openstack/swift: Fix .coveragrc to prevent nose tests error https://review.openstack.org/225591 | 10:42 |
| *** haomaiwa_ has quit IRC | 11:01 | |
| *** 6JTAB24P4 has joined #openstack-swift | 11:01 | |
| *** eranrom has joined #openstack-swift | 11:04 | |
| *** sanchitmalhotra has joined #openstack-swift | 11:21 | |
| *** sanchitmalhotra1 has quit IRC | 11:22 | |
| *** silor has joined #openstack-swift | 11:25 | |
| *** 6JTAB24P4 has quit IRC | 11:41 | |
| *** hezhiqiang has joined #openstack-swift | 11:44 | |
| *** jordanP has joined #openstack-swift | 11:52 | |
| *** hezhiqiang has quit IRC | 11:57 | |
| *** sanchitmalhotra1 has joined #openstack-swift | 12:00 | |
| *** sanchitmalhotra has quit IRC | 12:01 | |
| *** sanchitmalhotra has joined #openstack-swift | 12:06 | |
| *** sanchitmalhotra1 has quit IRC | 12:08 | |
| *** sanchitmalhotra1 has joined #openstack-swift | 12:16 | |
| *** thurloat_isgone is now known as thurloat | 12:17 | |
| *** sanchitmalhotra has quit IRC | 12:18 | |
| *** mahatic has joined #openstack-swift | 12:40 | |
| *** nadeem has joined #openstack-swift | 12:40 | |
| *** nadeem has quit IRC | 12:56 | |
| *** nadeem has joined #openstack-swift | 12:56 | |
| *** dustins has joined #openstack-swift | 13:05 | |
| *** kairo has joined #openstack-swift | 13:10 | |
| *** breitz has joined #openstack-swift | 13:16 | |
| *** amoturi has joined #openstack-swift | 13:21 | |
| *** haomaiwang has joined #openstack-swift | 13:23 | |
| *** hrou has joined #openstack-swift | 13:26 | |
| *** SkyRocknRoll has quit IRC | 13:28 | |
| *** barker has joined #openstack-swift | 13:36 | |
| *** jordanP has quit IRC | 13:50 | |
| *** vinsh has quit IRC | 13:51 | |
| *** pberis has joined #openstack-swift | 13:54 | |
| *** mc_nair has joined #openstack-swift | 13:56 | |
| *** haomaiwang has quit IRC | 14:01 | |
| *** david-ly_ is now known as david-lyle | 14:01 | |
| *** haomaiwang has joined #openstack-swift | 14:01 | |
| *** cazino has joined #openstack-swift | 14:02 | |
| *** wbhuber has joined #openstack-swift | 14:07 | |
| *** jordanP has joined #openstack-swift | 14:10 | |
| *** sanchitmalhotra1 has quit IRC | 14:10 | |
| *** pberis has quit IRC | 14:13 | |
| *** jordanP has quit IRC | 14:18 | |
| *** jordanP has joined #openstack-swift | 14:19 | |
| *** david-lyle has quit IRC | 14:22 | |
| *** david-lyle has joined #openstack-swift | 14:26 | |
| *** barker has quit IRC | 14:29 | |
| *** barker has joined #openstack-swift | 14:31 | |
| *** mahatic has quit IRC | 14:35 | |
| *** mahatic has joined #openstack-swift | 14:36 | |
| *** jlhinson has joined #openstack-swift | 14:39 | |
| *** clyps_ has quit IRC | 14:44 | |
| *** clyps_ has joined #openstack-swift | 14:44 | |
| *** ejat has joined #openstack-swift | 14:45 | |
| ejat | hi .. | 14:45 |
| ejat | i tried to run rsync rsync://pub@localhost/ | 14:45 |
| ejat | but no output | 14:45 |
| ejat | appreciate if someone can help me | 14:45 |
| *** wshao has joined #openstack-swift | 14:51 | |
| *** torgomatic has quit IRC | 14:55 | |
| *** torgomatic has joined #openstack-swift | 14:55 | |
| *** ChanServ sets mode: +v torgomatic | 14:55 | |
| *** Fin1te has joined #openstack-swift | 14:57 | |
| *** mahatic_ has joined #openstack-swift | 14:58 | |
| *** peluse_ has joined #openstack-swift | 14:58 | |
| *** peluse has quit IRC | 14:59 | |
| *** albertom has quit IRC | 14:59 | |
| *** minwoob has joined #openstack-swift | 14:59 | |
| *** mahatic has quit IRC | 15:00 | |
| *** haomaiwang has quit IRC | 15:01 | |
| *** haomaiwang has joined #openstack-swift | 15:01 | |
| *** devlaps has joined #openstack-swift | 15:02 | |
| *** barker has quit IRC | 15:03 | |
| *** albertom has joined #openstack-swift | 15:05 | |
| *** barker has joined #openstack-swift | 15:08 | |
| *** annegentle has joined #openstack-swift | 15:08 | |
| *** jkugel has joined #openstack-swift | 15:10 | |
| *** barker has quit IRC | 15:12 | |
| mahatic_ | clayg: this was the patch 223489 I was referring to the other day, for yoru review | 15:13 |
| patchbot | mahatic_: https://review.openstack.org/#/c/223489/ | 15:13 |
| mahatic_ | your* | 15:13 |
| *** silor1 has joined #openstack-swift | 15:19 | |
| *** devlaps has quit IRC | 15:19 | |
| *** dustins is now known as dustins|away | 15:23 | |
| *** silor has quit IRC | 15:23 | |
| *** silor1 is now known as silor | 15:23 | |
| *** silor has quit IRC | 15:25 | |
| *** silor has joined #openstack-swift | 15:25 | |
| *** devlaps has joined #openstack-swift | 15:28 | |
| clayg | mahatic_: thanks! | 15:30 |
| acoles | notmyname: i just updated patch https://review.openstack.org/#/c/225540/ with the coveragrc fix that broke gate, i forgot that before rechecking earlier | 15:32 |
| clayg | poor gate | 15:33 |
| *** pberis has joined #openstack-swift | 15:35 | |
| *** pberis1 has joined #openstack-swift | 15:36 | |
| *** itlinux has joined #openstack-swift | 15:39 | |
| *** pberis has quit IRC | 15:40 | |
| *** pberis1 is now known as pberis | 15:40 | |
| *** garthb has joined #openstack-swift | 15:47 | |
| *** nadeem has quit IRC | 15:51 | |
| *** mc_nair has quit IRC | 15:52 | |
| *** haomaiwang has quit IRC | 15:54 | |
| *** haomaiwang has joined #openstack-swift | 15:55 | |
| *** cazino has left #openstack-swift | 15:59 | |
| *** haomaiwang has quit IRC | 16:01 | |
| *** haomaiwang has joined #openstack-swift | 16:01 | |
| *** wshao has quit IRC | 16:01 | |
| *** gyee has joined #openstack-swift | 16:02 | |
| *** proteusguy_ has quit IRC | 16:02 | |
| *** alejandrito has joined #openstack-swift | 16:03 | |
| *** nadeem has joined #openstack-swift | 16:03 | |
| notmyname | acoles: thanks | 16:05 |
| *** breitz has quit IRC | 16:06 | |
| *** breitz has joined #openstack-swift | 16:09 | |
| *** proteusguy_ has joined #openstack-swift | 16:15 | |
| openstackgerrit | Kota Tsuyuzaki proposed openstack/swift: Fix missing container update https://review.openstack.org/186735 | 16:18 |
| *** zhill has joined #openstack-swift | 16:21 | |
| *** patchbot has quit IRC | 16:27 | |
| *** patchbot has joined #openstack-swift | 16:27 | |
| *** rledisez has quit IRC | 16:29 | |
| openstackgerrit | Merged openstack/swift-specs: Change ignore-errors to ignore_errors https://review.openstack.org/225835 | 16:33 |
| *** annegentle has quit IRC | 16:38 | |
| *** annegentle has joined #openstack-swift | 16:39 | |
| *** lyrrad has joined #openstack-swift | 16:39 | |
| *** dustins|away is now known as dustins | 16:45 | |
| *** chlong has quit IRC | 16:50 | |
| *** tongli has joined #openstack-swift | 16:55 | |
| *** haomaiwang has quit IRC | 16:58 | |
| *** haomaiwang has joined #openstack-swift | 16:59 | |
| *** thurloat is now known as thurloat_isgone | 17:00 | |
| *** haomaiwang has quit IRC | 17:01 | |
| *** haomaiwang has joined #openstack-swift | 17:01 | |
| *** haomaiwang has quit IRC | 17:01 | |
| *** chlong has joined #openstack-swift | 17:02 | |
| *** haomaiwang has joined #openstack-swift | 17:02 | |
| *** haomaiwang has quit IRC | 17:03 | |
| *** haomaiwang has joined #openstack-swift | 17:03 | |
| *** Fin1te has quit IRC | 17:06 | |
| *** haomaiwang has quit IRC | 17:11 | |
| *** devlaps has quit IRC | 17:12 | |
| *** itlinux has quit IRC | 17:13 | |
| *** jordanP has quit IRC | 17:15 | |
| clayg | acoles: hey - patch 181407 - seems reasonable in principle | 17:15 |
| patchbot | clayg: https://review.openstack.org/#/c/181407/ | 17:15 |
| clayg | acoles: I'm not sure if it totally jives with the multi-frag on primary behavior? | 17:15 |
| clayg | acoles: patch 174875 too | 17:16 |
| patchbot | clayg: https://review.openstack.org/#/c/174875/ | 17:16 |
| acoles | clayg: hi. wow those are old. i think there is stuff in them that would be good to do. idk if its a priority for liberty. | 17:19 |
| acoles | 181407 - would need a bit or rework to allow multi frags on primaries | 17:20 |
| acoles | the goodness in it is not trying to send frags to handoffs that will conflict | 17:20 |
| clayg | acoles: right! seems totally reasonable! | 17:21 |
| acoles | iirc 174875 tries to be nicer when that ^^ does happen, so the whole ssync transaction is not aborted due to one frag conflict | 17:21 |
| clayg | acoles: right - per object sync error reporting or something - it was genius! | 17:22 |
| acoles | clayg: right now i'm in down huge rabbit hole with optimistic gets - i can see the light and it'll be great but its taking me hours | 17:22 |
| acoles | well, days :) | 17:22 |
| clayg | acoles: k, be sure to push what you have before you sign off | 17:23 |
| clayg | acoles: thanks for working on it | 17:23 |
| acoles | clayg: k | 17:23 |
| clayg | acoles: I think the diskfile stuff is done - i need to look at the object server stuff - are you up in the proxy in the next change now? | 17:23 |
| acoles | clayg: oh its "fun" it that mind-bending way we seem to be attracted to | 17:24 |
| clayg | nom nom nom | 17:25 |
| acoles | clayg: i'm working on 215276. yes in proxy obj.py. but i have had to rework the diskfile gather-er stuff so that i can report alternative 'durable' stuff to the proxy for it be smarter about where to go try again. | 17:27 |
| acoles | clayg: actually i think i may have ended up implementing something you commented wrt to sorting all the disk files etc etc | 17:27 |
| acoles | so patch 215276 on gerrit sorta works but is horribly inefficient and has some corner case where it wont find an object. | 17:28 |
| patchbot | acoles: https://review.openstack.org/#/c/215276/ | 17:28 |
| acoles | clayg: where i am heading will be much smarter in proxy and will probably end up doing peluse's alt frags thing as a side effect | 17:29 |
| *** aix has quit IRC | 17:29 | |
| *** mac_ified has joined #openstack-swift | 17:30 | |
| acoles | notmyname: what's our deadline for release? as in what used to be RC | 17:32 |
| *** subscope has joined #openstack-swift | 17:37 | |
| notmyname | the end of next week is the goal | 17:39 |
| notmyname | AAAAAAAAAAAAAHHHHH! | 17:39 |
| clayg | nice | 17:39 |
| *** silor has quit IRC | 17:40 | |
| *** klrmn1 has joined #openstack-swift | 17:41 | |
| *** amoturi has left #openstack-swift | 17:50 | |
| clayg | acoles: so everything obj/proxy is going to go into patch 215276? | 17:52 |
| patchbot | clayg: https://review.openstack.org/#/c/215276/ | 17:52 |
| acoles | notmyname: 10/2 right? | 17:53 |
| notmyname | acoles: yeah. the secret is that there isn't a hard date. but we need something around then | 17:54 |
| acoles | clayg: at the moment that was my plan because the proxy and diskfile changes are coupled. like i said i have had to change the diskfile stuff i had in 215276. but i'm open to change that plan. | 17:55 |
| *** geaaru has quit IRC | 17:56 | |
| acoles | notmyname: so 10/2ish :) | 17:56 |
| *** mc_nair has joined #openstack-swift | 17:56 | |
| acoles | notmyname: i feared end this week so thats cool | 17:56 |
| notmyname | acoles: good news! ;-) | 17:57 |
| acoles | i can kick back spend more time pressing make-everything-ok.com now. i'm sure its gonna work. | 17:57 |
| *** Fin1te has joined #openstack-swift | 18:02 | |
| *** chsc has joined #openstack-swift | 18:03 | |
| dhellmann | notmyname: I'm planning to use swiftclient 2.6.0 as the basis for your stable/liberty branch, does that work for you? | 18:04 |
| notmyname | dhellmann: yes | 18:05 |
| *** mahatic_ has quit IRC | 18:17 | |
| *** jrichli has joined #openstack-swift | 18:24 | |
| acoles | clayg: your comment on patch 215276 version 5 about sorting listdir etc. i think thats what i have ended up doing ;) | 18:30 |
| patchbot | acoles: https://review.openstack.org/#/c/215276/ | 18:30 |
| clayg | acoles: you're just trying to spread the blame again - but if you wrote it i'm sure it's good | 18:30 |
| clayg | acoles: ah yeah - i'm looking at that patch - so far I like what I see in diskfile - so KUDOS! | 18:31 |
| acoles | clayg: heh. | 18:31 |
| acoles | clayg: wait. i didn't push that yet! | 18:31 |
| acoles | did i ? | 18:31 |
| acoles | clayg: i am so confused by myself! | 18:32 |
| clayg | WAT? | 18:33 |
| peluse_ | FYI ran 4 variations of EC parms (segment size, chunk sizes) for perf tests and will have powerpoint available here in a few hrs, will email to those I know are interested unless someone knows where to post a powerpoint | 18:33 |
| clayg | OH VERSION 5 | 18:33 |
| acoles | clayg: no patch set 7 i just did the cleanups you suggested. next patchset (not yet there ) will be different. | 18:33 |
| clayg | oh... well shit - maybe I should work on something else | 18:34 |
| clayg | HOW COULD IT POSSIBLY GET ANY BETTER!? | 18:34 |
| * peluse_ has to go look now, this can't be real! | 18:34 | |
| clayg | peluse_: I normally just throw it in google drive and make it public | 18:34 |
| peluse_ | clayg, oh yeha, I guess I'll do that and post a thingy here when its up there | 18:35 |
| clayg | peluse_: no don't do it - acoles says it spoil your dinner | 18:35 |
| peluse_ | i haven't even had lunch yet?? | 18:35 |
| acoles | clayg: yeah well my concern is i'm making it worse but we'll see. trying to get something coherent to push tonight. | 18:35 |
| *** peluse_ is now known as peluse | 18:35 | |
| *** ChanServ sets mode: +v peluse | 18:35 | |
| *** eranrom has quit IRC | 18:46 | |
| *** cdelatte has joined #openstack-swift | 18:50 | |
| *** nadeem has quit IRC | 18:56 | |
| *** annegentle has quit IRC | 19:03 | |
| *** jrichli has quit IRC | 19:04 | |
| *** nadeem has joined #openstack-swift | 19:07 | |
| *** jrichli has joined #openstack-swift | 19:10 | |
| *** potato_farmer has joined #openstack-swift | 19:12 | |
| *** potato_farmer has left #openstack-swift | 19:19 | |
| *** mc_nair has quit IRC | 19:25 | |
| *** itlinux has joined #openstack-swift | 19:26 | |
| *** jrichli_ has joined #openstack-swift | 19:27 | |
| *** jrichli has quit IRC | 19:27 | |
| *** breitz has quit IRC | 19:35 | |
| *** breitz has joined #openstack-swift | 19:37 | |
| openstackgerrit | Alistair Coles proposed openstack/swift: diskfile: make get_ondisk_files return a dict https://review.openstack.org/222706 | 19:40 |
| openstackgerrit | Alistair Coles proposed openstack/swift: Enable object server to return non-durable data https://review.openstack.org/215276 | 19:40 |
| *** nadeem has quit IRC | 19:41 | |
| openstackgerrit | Janie Richling proposed openstack/swift: Encrypting/Decrypting object metadata https://review.openstack.org/203310 | 19:45 |
| clayg | acoles: oh my goodness i'm so excited! | 19:46 |
| acoles | clayg: peluse just pushed my latest on patch 215276 before i leave, but it does need more work. this version uses less requests when the optimistic get doesn't produce the goods. | 19:47 |
| patchbot | acoles: https://review.openstack.org/#/c/215276/ | 19:47 |
| acoles | clayg: if you do have chance to review could you tell me a better to do the NodeIter alt nodes injection thing? | 19:48 |
| acoles | clayg: oh, and i know there are tests missing. WIP really means WIP. | 19:48 |
| *** DericHorn-HP has joined #openstack-swift | 19:48 | |
| clayg | acoles: yup - got it - get outta there | 19:49 |
| peluse | acoles, oh yeah, blame it on me :) | 19:49 |
| peluse | so now I'm curious. on patch 222706 acoles added set 3 after set 2 was approved but set 2 didn't get merged because of an unrelated bug recheck. I assume changes from set 3 need to manually broken out as a new patch? or, once the tests are completed I wonder if it will be smart enough to drop the +A and leav e set 3 there to be reviewed?? | 19:52 |
| patchbot | peluse: https://review.openstack.org/#/c/222706/ | 19:52 |
| *** amoturi has joined #openstack-swift | 19:53 | |
| acoles | peluse: yeah i blew it out of zuul i think | 19:53 |
| peluse | acoles, hope that's what happens | 19:54 |
| acoles | peluse: i just re-added +A - i rebased my dependent on master so it got the gate fix (coverargc thing) and 222706 got rebased too, but there is no material change to it. | 19:54 |
| peluse | ahhh | 19:54 |
| peluse | what fun! | 19:54 |
| peluse | +6 | 19:55 |
| acoles | good night | 19:56 |
| peluse | later | 19:57 |
| *** acoles is now known as acoles_ | 19:58 | |
| clayg | man, so i was going to start rebasing all the patches - but now paul send out pretty pictures! | 20:01 |
| peluse | looks like google drive on my mac wants me to go home to post the perf stuff (doesnt like proxy) so darn in, guess I will do that and post from there | 20:04 |
| *** tongli has quit IRC | 20:09 | |
| *** annegentle has joined #openstack-swift | 20:10 | |
| clayg | peluse: looks super interesting at first glance it seems like we've gotten validation on the increased througput on large writes - and sort of validated the small op req/s trade off - nice work! | 20:11 |
| *** amoturi has left #openstack-swift | 20:12 | |
| notmyname | on the other hand, it's interesting to me that you aren't seeing better performance with larger objects (since less data is being written to disks in EC than in repl). maybe I'm reading it wrong, though | 20:17 |
| clayg | notmyname: you can see it on the op/s graphs down at the bottom 32M and 64M writes | 20:17 |
| *** itlinux has quit IRC | 20:22 | |
| *** jrichli_ has quit IRC | 20:30 | |
| *** jrichli has joined #openstack-swift | 20:35 | |
| openstackgerrit | Janie Richling proposed openstack/swift: Change fake_footers to replace response headers instead of append https://review.openstack.org/222953 | 20:36 |
| *** trifon has quit IRC | 20:37 | |
| *** acoles_ has quit IRC | 20:39 | |
| *** jrichli has quit IRC | 20:44 | |
| *** Fin1te has quit IRC | 20:46 | |
| *** acoles_ has joined #openstack-swift | 20:48 | |
| *** ChanServ sets mode: +v acoles_ | 20:48 | |
| *** acoles_ is now known as acoles | 20:48 | |
| *** aix has joined #openstack-swift | 20:57 | |
| *** dustins has quit IRC | 21:01 | |
| *** itlinux has joined #openstack-swift | 21:02 | |
| *** Fin1te has joined #openstack-swift | 21:04 | |
| *** agomez has joined #openstack-swift | 21:07 | |
| agomez | Any reason why a cluster with 3x replica would show only 2 primary replicas in get nodes ? | 21:11 |
| agomez | for some partitions | 21:12 |
| clayg | :'( | 21:12 |
| clayg | agomez: https://bugs.launchpad.net/swift/+bug/1452431 maybe :\ | 21:12 |
| openstack | Launchpad bug 1452431 in OpenStack Object Storage (swift) "some parts replicas assigned to duplicate devices in the ring" [Critical,Confirmed] - Assigned to Samuel Merritt (torgomatic) | 21:12 |
| agomez | though my repica count is less than devices | 21:14 |
| agomez | root@ctl-bfy-1:~/test-ring# swift-ring-builder container.builder | head | 21:14 |
| agomez | container.builder, build version 532 | 21:14 |
| agomez | 32768 partitions, 3.000000 replicas, 1 regions, 1 zones, 82 devices, 1.47 balance | 21:14 |
| agomez | The minimum number of hours before a partition can be reassigned is 0 | 21:14 |
| agomez | swift-get-nodes container.ring.gz -p 5252 | 21:14 |
| agomez | Account None | 21:14 |
| agomez | ContainerNone | 21:14 |
| agomez | Object None | 21:14 |
| agomez | Partition5252 | 21:14 |
| agomez | Hash None | 21:14 |
| agomez | Server:Port Device10.40.140.15:6001 disk-0 | 21:14 |
| agomez | Server:Port Device10.40.140.17:6001 disk-1 | 21:14 |
| agomez | Server:Port Device10.40.140.16:6001 disk-1 [Handoff] | 21:14 |
| agomez | Server:Port Device10.40.140.9:6001 disk-7 [Handoff] | 21:14 |
| agomez | curl -I -XHEAD "http://10.40.140.15:6001/disk-0/5252/None" | 21:14 |
| agomez | curl -I -XHEAD "http://10.40.140.17:6001/disk-1/5252/None" | 21:14 |
| clayg | oh my :'( | 21:14 |
| agomez | curl -I -XHEAD "http://10.40.140.16:6001/disk-1/5252/None" # [Handoff] | 21:14 |
| agomez | curl -I -XHEAD "http://10.40.140.9:6001/disk-7/5252/None" # [Handoff] | 21:14 |
| clayg | 82 devices and 1.47 balance - seems odd - are the devices lots of different sizes? | 21:15 |
| agomez | I just weight to a new device | 21:15 |
| agomez | added weight* | 21:15 |
| agomez | the weight on the new device is greater than the others | 21:16 |
| agomez | it is a bit bigger | 21:16 |
| *** Fin1te has quit IRC | 21:18 | |
| clayg | ok, well those no good for reason for it to do that - so it's probably a bug | 21:21 |
| agomez | do you think if I rebalance again it will correct itself ? | 21:21 |
| clayg | you could know for sure if it's that bug i referenced by running the validate script that's attached to it against your ring | 21:21 |
| clayg | agomez: depends - you could make a backup/copy and run rebalance with --debug and post the output | 21:22 |
| agomez | ok | 21:22 |
| clayg | if the validate script on lp bug #1452431 fails on your ring - the --debug output would be useful to have on that bug | 21:22 |
| openstack | Launchpad bug 1452431 in OpenStack Object Storage (swift) "some parts replicas assigned to duplicate devices in the ring" [Critical,Confirmed] https://launchpad.net/bugs/1452431 - Assigned to Samuel Merritt (torgomatic) | 21:22 |
| agomez | am working on backup copy now to test just waiting on min part hours | 21:22 |
| clayg | there's a command you can give ring-builder called pretend_min_part_hours_passed | 21:23 |
| clayg | don't ever use it on a real builder | 21:23 |
| agomez | root@ctl-bfy-1:~/test-ring# ./validate.py container.builder | 21:24 |
| agomez | set([2, 3]) 3.0 | 21:24 |
| agomez | ERROR: some parts have different numbers of device ids for their replicas! | 21:24 |
| clayg | i hate that stupid bug :'( | 21:25 |
| agomez | so I am being affected by this bug? | 21:26 |
| agomez | what does this even mean ? | 21:26 |
| clayg | yes | 21:26 |
| clayg | it means you have bad rings - bad is pretty subjective tho - there's some set of inputs to ring-builder that produce this crappy behavior - but we don't really understand the constraints on the input or the cause of the beavhior - but we know it's bad | 21:27 |
| clayg | i say "we" - i mean "me" - torgomatic and cschwede & maybe timburke could explain the problem better than I could | 21:27 |
| agomez | do I simply need many more devices ? | 21:29 |
| clayg | agomez: i'm not entirely sure - most of the time it seems to have more to do with the relative weights of the failure domains - like you have one server that's got 2x more that some otherone or something | 21:30 |
| agomez | hmm | 21:30 |
| clayg | bbiab | 21:31 |
| MooingLemur | Just updated from 2.2.0 to 2.4.0 on my QA cluster, and suddenly account-replicator and container-replicator are not seeing any containers to check. | 21:32 |
| MooingLemur | the rings havent actually changed in a couple of years... was there a drop of the old ring format? | 21:33 |
| MooingLemur | (or perhaps a regression) | 21:33 |
| *** hrou has quit IRC | 21:41 | |
| minwoob | peluse: How did you decide that object_size < segment_size is the appropriate threshold for preferring replication over EC? | 21:42 |
| agomez | root@ctl-bfy-1:~/test-ring# swift-ring-builder container.builder set_weight 10.40.140.15 300 | 21:43 |
| agomez | d81r1z1-10.40.140.15:6001R10.40.140.15:6001/disk-0_"" weight set to 300.0 | 21:43 |
| agomez | root@ctl-bfy-1:~/test-ring# swift-ring-builder container.builder rebalance | 21:43 |
| agomez | Reassigned 370 (1.13%) partitions. Balance is now 0.05. | 21:43 |
| agomez | root@ctl-bfy-1:~/test-ring# ./validate.py container.builder | 21:43 |
| agomez | SUCCESS: all parts have 3 device ids! | 21:43 |
| minwoob | peluse: And have you had a chance to consider how the particular (n_data + n_parity) scheme changes the threshold? | 21:44 |
| clayg | MooingLemur: I can't think of anyhting that would have changed to cause that | 21:47 |
| clayg | oh poppy is CDNaaS - get it "pop" - too funny | 21:50 |
| * briancline has always wondered about that | 21:54 | |
| *** jlhinson has quit IRC | 22:00 | |
| *** alejandrito has quit IRC | 22:05 | |
| peluse | minwoob, did you see the .ppt? | 22:07 |
| *** breitz has quit IRC | 22:08 | |
| *** breitz has joined #openstack-swift | 22:08 | |
| peluse | minwoob, but to answer the question (both actually) we just went with matching the size of a fragment to that of the network/disk chunk defaults. For a 10:14 that's 64K * 10 so we used 640K segment | 22:09 |
| minwoob | peluse: Could you send me the ppt? | 22:09 |
| peluse | minwoob, haven't changed anything other than what's in the deck yet which I can't seem to post because google drive isn't working for me but one would assume that if you take the proxy chunk size * n_data you'd probably have a good match | 22:10 |
| peluse | minwoob, sure send me your email | 22:10 |
| minwoob | peluse: minwoob@us.ibm.com | 22:10 |
| peluse | so any other mac os people, I installed google drive and it says its running but I don't see anything, anywhere... | 22:10 |
| peluse | minwoob, just sent it you you and huber both | 22:11 |
| minwoob | peluse: Also including wbhuber@us.ibm.com would be good. | 22:11 |
| minwoob | peluse: Thanks. | 22:11 |
| peluse | heh, beat you to it! | 22:11 |
| minwoob | You read my mind. | 22:12 |
| wbhuber | peluse: minwoob: thank you both! | 22:12 |
| briancline | summit related question... anyone know if Honolulu is a decent place to re-enter the country? i.e. customs lines not taking 4 hours | 22:13 |
| peluse | briancline, I've done it many times, not recently, and it sucked | 22:14 |
| *** jkugel has quit IRC | 22:14 | |
| briancline | :| | 22:15 |
| *** otoolee has quit IRC | 22:15 | |
| peluse | so there's some info that I didn't put in the perf preso yet Im sure, feel free to ask questions. I was using cosbnech w/4 workload driving machines for a total of 512 worker threads for one thing | 22:15 |
| briancline | trying to avoid what happened from paris -> chicago last year. customs was awful, which you wouldn't think it'd be at o'hare | 22:15 |
| briancline | probably inevitable though | 22:17 |
| *** itlinux has quit IRC | 22:19 | |
| *** itlinux has joined #openstack-swift | 22:33 | |
| *** pberis has quit IRC | 22:36 | |
| *** pberis has joined #openstack-swift | 22:37 | |
| *** peluse_ has joined #openstack-swift | 22:37 | |
| peluse_ | initial EC perf data: https://drive.google.com/a/asu.edu/file/d/0Bw_XIyQ62S1vSWs4QWJ5d19aLWM/view?usp=sharing | 22:38 |
| *** annegentle has quit IRC | 22:39 | |
| *** peluse__ has joined #openstack-swift | 22:42 | |
| peluse | if someone can tell me if that link works or not that would be great :) | 22:42 |
| *** pberis has quit IRC | 22:42 | |
| *** peluse_ has quit IRC | 22:44 | |
| *** peluse is now known as peluse2 | 22:45 | |
| *** peluse__ is now known as peluse | 22:45 | |
| *** ChanServ sets mode: +v peluse | 22:46 | |
| mattoliverau | Morning | 22:48 |
| peluse | morning | 22:48 |
| mattoliverau | peluse: need to request access | 22:48 |
| mattoliverau | So not quite working | 22:49 |
| peluse | mattoliverau: hmmm... OK not sure what I need to do but I'll look again | 22:49 |
| *** cdelatte has quit IRC | 22:49 | |
| peluse2 | mattoliverau, oh I see. I shared with only people at ASU :) One min | 22:50 |
| peluse | OK, please try this real quick (and thanks) https://drive.google.com/file/d/0Bw_XIyQ62S1vSWs4QWJ5d19aLWM/view?usp=sharing | 22:51 |
| *** proteusguy__ has joined #openstack-swift | 22:53 | |
| wbhuber | peluse: that works! | 22:54 |
| *** proteusguy_ has quit IRC | 22:55 | |
| peluse | wbhuber: cool, thanks | 22:58 |
| *** aix has quit IRC | 23:02 | |
| *** agomez has quit IRC | 23:07 | |
| openstackgerrit | Merged openstack/swift: Merge branch 'master' into feature/crypto https://review.openstack.org/225540 | 23:08 |
| *** DericHorn-HP has quit IRC | 23:09 | |
| *** itlinux has quit IRC | 23:14 | |
| *** itlinux has joined #openstack-swift | 23:17 | |
| *** chsc has quit IRC | 23:19 | |
| *** jamielennox|away is now known as jamielennox | 23:23 | |
| *** km has joined #openstack-swift | 23:25 | |
| openstackgerrit | Merged openstack/python-swiftclient: Change ignore-errors to ignore_errors https://review.openstack.org/225826 | 23:28 |
| *** peluse has quit IRC | 23:37 | |
| *** zhill has quit IRC | 23:39 | |
| *** itlinux has quit IRC | 23:43 | |
| *** minwoob has quit IRC | 23:57 | |
| *** kei_yama has joined #openstack-swift | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!