| openstackgerrit | Samuel Merritt proposed a change to openstack/swift: Fix SLO test with old simplejson https://review.openstack.org/99805 | 00:06 |
|---|---|---|
| torgomatic | clayg: so after I s/json.JSONDecodeError/ValueError/g, the probetests are all passing for me (one skip, not sure what it is) | 00:16 |
| *** matsuhashi has joined #openstack-swift | 00:19 | |
| portante | torgomatic: did you look at the stack trace I posted above? | 00:26 |
| torgomatic | portante: nope | 00:27 |
| portante | the one where file not found for /tmp? | 00:27 |
| torgomatic | yeah, have you got /srv/1/node/sdb1/tmp/ ? | 00:27 |
| portante | ls -ld /srv/*/node/*/tmp: No such file or directory | 00:29 |
| portante | but all tmp-1 exist | 00:29 |
| torgomatic | wacky; isn't the object server supposed to create those if it needs to? | 00:29 |
| torgomatic | oh, so it's using the wrong tempdir | 00:29 |
| portante | yes | 00:29 |
| torgomatic | welp | 00:29 |
| portante | the code has a hard-coded 0 in async_update() call in delete_at I believe | 00:30 |
| portante | but I was not sure I understood the code right | 00:30 |
| portante | the code used to not care what tmpdir it uses | 00:31 |
| torgomatic | yeah, it used to share quarantine and tmp dirs between all policies, but we changed that | 00:34 |
| torgomatic | the quarantine one is certainly a good idea IMO; if stuff's mistakenly quarantined, I want to know where to put it back to | 00:34 |
| torgomatic | the tmp one, I dunno. I have no strong opinion either way. | 00:34 |
| clayg | torgomatic: well either way, the change that you and peluse_ worked on didn't do the get_policy_dir bits for the tmp dir and it only showed up for me in probetests | 00:36 |
| clayg | portante: but even in the traceback you pasted it looked like the call to get_policy_dir was in there, unless i'm mis-reading? | 00:36 |
| clayg | torgomatic: you skip is probably container-sync - and you should set that up | 00:36 |
| torgomatic | I don't see any failures, but maybe that's because I have a policy 0 defined | 00:36 |
| clayg | torgomatic: i thought portante did as well? i'm assuming he'd try to get something working before he tried to break it - dunno if he fixed the 2x ring tho - peluse_ confirmed that's not gunna work | 00:37 |
| *** haomaiwang has joined #openstack-swift | 00:38 | |
| portante | clayg: yes, but delete_at appears to hard-code a 0 in the call to async_update() | 00:39 |
| portante | clayg, torgomatic: i have the storage policies defined as in the SAIO, so a 0 and a 1 | 00:40 |
| torgomatic | portante: hm, maybe just bad luck for you then? something something test ordering? | 00:40 |
| portante | I have resetswift multiple times and get the same thing though | 00:40 |
| portante | something else in my environment that is leaking into the tests? | 00:41 |
| portante | fukshima? | 00:41 |
| portante | something I ate? | 00:41 |
| torgomatic | sure, maybe it's the order that the test files come out of readdir() so they run in a different order than on my box | 00:41 |
| torgomatic | dunno | 00:41 |
| portante | torgomatic: I though the tests were sorted? | 00:43 |
| portante | by name that is | 00:43 |
| * torgomatic is not a nose expert | 00:43 | |
| *** dmorita has joined #openstack-swift | 01:16 | |
| *** nosnos has joined #openstack-swift | 01:21 | |
| *** NM has joined #openstack-swift | 01:29 | |
| *** NM has quit IRC | 01:35 | |
| *** mkollaro has quit IRC | 01:39 | |
| *** haomaiwang has quit IRC | 01:46 | |
| *** haomaiwa_ has joined #openstack-swift | 01:53 | |
| *** kajinamit has joined #openstack-swift | 01:54 | |
| *** haomaiwa_ has quit IRC | 01:57 | |
| clayg | portante: yeah system accounts are always policy 0, hell most of them don't even have any "real" objects so it doesn't matter - there's not a comment that says that? | 02:00 |
| clayg | portante: no, not sorted by name, nose is all pretty random like that | 02:04 |
| clayg | portante: but the tests themselves shouldn't care the order they're run in :) | 02:04 |
| clayg | portante: oh oh oh, the async pickle for the delete-at - christ, yeah i'll try to track that down - nice catch! | 02:05 |
| *** nsquare has quit IRC | 02:19 | |
| clayg | acoles: so I think the fixed with hex offset will work fine for the content-type updates, but I agree that the offset there should be based on the x-timestamp of the POST | 02:26 |
| clayg | acoles: you just need to convert the x-timestamp of the post to an integer and then we need to pick a maximum width to maintain string order sorting | 02:28 |
| clayg | acoles: but I think storing the timestamp offset string as a decimal is wasteful of our bits when we have hex just *right there* ;) | 02:31 |
| openstackgerrit | Kota Tsuyuzaki proposed a change to openstack/swift: Efficient Replication for Distributed Region https://review.openstack.org/99824 | 02:39 |
| openstackgerrit | Kota Tsuyuzaki proposed a change to openstack/swift: Efficient Replication for Distributed Regions https://review.openstack.org/99824 | 02:44 |
| *** zhiyan_ is now known as zhiyan | 02:45 | |
| *** dmorita has quit IRC | 02:59 | |
| clayg | acoles: if you do int(post.ts) - int(data.ts) you get about 136 years out of 8 hex digits - but you loose metadata update resolution below a second which seems sorta crappy | 03:01 |
| clayg | acoles: if you take resolution out to a centi-microsecond (which seems not terrible) you only get like a month to do a POST before you have to overflow into the datafile timestamp - which is cirtainly long enough that you're outside of the consitency window and overflowing into the datafile timestamp would probably still end up working out as correct, but you loose the transparency of the update | 03:04 |
| clayg | OTOH if we take another 4 bytes and store 12 digit offsets you could have deca-microsecond resolution for 89 years before overflowing into the datafile timestamp and 10 times that if you go with centi-microseconds | 03:06 |
| clayg | so i'm inclined to bump the width on the resolution out to 12 digts - but I think we should leave the math up to the update code, and go with fixed with hex for the data format | 03:09 |
| * clayg briefly ponders an integer serialization encoding that would still be string sortable and uses more ascii chars | 03:10 | |
| *** nosnos has quit IRC | 03:16 | |
| *** gyee has quit IRC | 03:19 | |
| clayg | acoles: so if we go with a 64 digit base instead of 16, we could near century deca-microsecond resolution out of 8 digts - i recommend an alphabet of digits + letters + '$&' :D | 03:35 |
| *** nosnos has joined #openstack-swift | 03:52 | |
| clayg | oh nm, & is a terrible character to copy and paste around - maybe % | 03:53 |
| clayg | ... or we could just use 12 hex digts | 03:53 |
| *** openfly has quit IRC | 03:59 | |
| *** matsuhas_ has joined #openstack-swift | 04:03 | |
| *** matsuhas_ has quit IRC | 04:04 | |
| *** matsuhas_ has joined #openstack-swift | 04:05 | |
| *** matsuhashi has quit IRC | 04:05 | |
| *** matsuhas_ has quit IRC | 04:12 | |
| *** matsuhashi has joined #openstack-swift | 04:13 | |
| *** madhuri has quit IRC | 04:16 | |
| *** haomaiwa_ has joined #openstack-swift | 04:31 | |
| *** haomaiw__ has joined #openstack-swift | 04:33 | |
| *** haomaiwa_ has quit IRC | 04:37 | |
| *** psharma has joined #openstack-swift | 04:41 | |
| *** kashyapk has joined #openstack-swift | 04:41 | |
| openstackgerrit | A change was merged to openstack/swift: Add debug logging to container sync requests https://review.openstack.org/95911 | 04:47 |
| *** zhiyan is now known as zhiyan_ | 05:06 | |
| *** ppai has joined #openstack-swift | 05:08 | |
| *** kashyapk has quit IRC | 05:10 | |
| *** kashyapk has joined #openstack-swift | 05:10 | |
| *** zhiyan_ is now known as zhiyan | 05:13 | |
| *** mkollaro has joined #openstack-swift | 05:32 | |
| *** nthacker_ has joined #openstack-swift | 05:39 | |
| *** nthacker has quit IRC | 05:42 | |
| *** kashyapk has quit IRC | 05:43 | |
| *** kashyapk has joined #openstack-swift | 05:44 | |
| *** zaitcev has quit IRC | 05:49 | |
| *** matsuhashi has quit IRC | 05:52 | |
| *** matsuhashi has joined #openstack-swift | 05:57 | |
| *** openfly has joined #openstack-swift | 05:57 | |
| *** zhiyan is now known as zhiyan_ | 06:32 | |
| *** jamiehannaford has joined #openstack-swift | 06:48 | |
| *** jamiehan_ has joined #openstack-swift | 06:52 | |
| *** jamiehannaford has quit IRC | 06:54 | |
| *** zhiyan_ is now known as zhiyan | 07:00 | |
| *** kashyapk has quit IRC | 07:01 | |
| *** kashyapk has joined #openstack-swift | 07:02 | |
| *** kashyapk has quit IRC | 07:06 | |
| *** jamiehan_ has quit IRC | 07:11 | |
| *** jamiehannaford has joined #openstack-swift | 07:12 | |
| *** jamiehan_ has joined #openstack-swift | 07:21 | |
| *** haomaiw__ has quit IRC | 07:22 | |
| *** haomaiwang has joined #openstack-swift | 07:23 | |
| *** jamiehannaford has quit IRC | 07:24 | |
| *** haomaiwang has quit IRC | 07:27 | |
| *** jamiehan_ has quit IRC | 07:29 | |
| *** jamiehannaford has joined #openstack-swift | 07:30 | |
| *** mkollaro has quit IRC | 07:35 | |
| *** matsuhas_ has joined #openstack-swift | 07:36 | |
| *** matsuhashi has quit IRC | 07:36 | |
| *** jamiehan_ has joined #openstack-swift | 07:39 | |
| *** jamiehannaford has quit IRC | 07:39 | |
| *** kashyapk has joined #openstack-swift | 07:41 | |
| *** zhiyan is now known as zhiyan_ | 08:10 | |
| *** zhiyan_ is now known as zhiyan | 08:30 | |
| *** haomaiwa_ has joined #openstack-swift | 08:34 | |
| *** haomaiw__ has joined #openstack-swift | 08:37 | |
| *** haomaiwa_ has quit IRC | 08:38 | |
| *** haomaiw__ has quit IRC | 08:41 | |
| *** haomaiwa_ has joined #openstack-swift | 08:42 | |
| *** kashyapk has quit IRC | 08:43 | |
| *** kashyapk has joined #openstack-swift | 08:43 | |
| *** zhiyan is now known as zhiyan_ | 08:52 | |
| *** mmcardle has joined #openstack-swift | 08:52 | |
| *** matsuhas_ has quit IRC | 08:57 | |
| *** zhiyan_ is now known as zhiyan | 09:00 | |
| *** zhiyan is now known as zhiyan_ | 09:09 | |
| *** d0ugal has quit IRC | 09:09 | |
| *** d0ugal has joined #openstack-swift | 09:10 | |
| *** zhiyan_ is now known as zhiyan | 09:10 | |
| *** matsuhashi has joined #openstack-swift | 09:12 | |
| openstackgerrit | Matthew Oliver proposed a change to openstack/swift: When a filesystem does't support xattr return a 507 https://review.openstack.org/99883 | 09:21 |
| mattoliverau | Well I worked late, time to call it a day. Have a great weekend all. notmyname I hope your recovering well. | 09:26 |
| *** mkerrin has quit IRC | 09:26 | |
| *** mkerrin has joined #openstack-swift | 09:33 | |
| *** matsuhashi has quit IRC | 09:43 | |
| *** kajinamit has quit IRC | 09:49 | |
| *** itarchitectkev has joined #openstack-swift | 09:51 | |
| *** haomaiwa_ has quit IRC | 09:52 | |
| itarchitectkev | Hey folks. An AWS announcement about S3 and encryption came my way today and naturally I'm getting asked about Swift's capabilities/developments in at-rest encryption. I saw Mirantis taking a stab at this a while back, where's Swift's stance on this? | 09:52 |
| *** haomaiwa_ has joined #openstack-swift | 09:54 | |
| itarchitectkev | (noting that encryption in S3 has been there for a while of course!) | 09:55 |
| *** ccorrigan has quit IRC | 09:57 | |
| ahale | imo its pointless to do encryption like that, if you want to encrypt data do it before giving it to a third party | 09:57 |
| ahale | but i heard barbican project believe otherwise and were working on something, or something | 09:58 |
| ahale | dunno what the actual swift stance is though | 09:58 |
| itarchitectkev | yeah, traditionally doing it in client before uploading has been ok, some folks seem to want this more transparent. users, who'd have them eh? | 10:03 |
| ahale | i guess if they care more about a checkbox on a security list and less about the reality then.. its a great idea | 10:03 |
| *** zhiyan is now known as zhiyan_ | 10:06 | |
| *** haomai___ has joined #openstack-swift | 10:06 | |
| itarchitectkev | It's more than that. Policy enforcement can be done server side, and not relying on a client. There are lots of reasons for this. | 10:06 |
| *** omame has joined #openstack-swift | 10:06 | |
| *** haomaiwa_ has quit IRC | 10:09 | |
| *** NM has joined #openstack-swift | 10:15 | |
| *** mkollaro has joined #openstack-swift | 10:15 | |
| *** haomai___ has quit IRC | 10:21 | |
| *** haomaiwang has joined #openstack-swift | 10:21 | |
| *** NM has quit IRC | 10:22 | |
| *** haomaiwang has quit IRC | 10:22 | |
| *** haomaiwa_ has joined #openstack-swift | 10:23 | |
| *** haomaiw__ has joined #openstack-swift | 10:24 | |
| *** haomaiw__ has quit IRC | 10:25 | |
| *** kashyapk has quit IRC | 10:25 | |
| *** haomaiwa_ has quit IRC | 10:25 | |
| *** haomaiwa_ has joined #openstack-swift | 10:26 | |
| *** nosnos has quit IRC | 10:26 | |
| *** kashyapk has joined #openstack-swift | 10:26 | |
| *** Longgeek_ has joined #openstack-swift | 10:27 | |
| openstackgerrit | Christian Schwede proposed a change to openstack/swift: Fix object auditor recon and logging https://review.openstack.org/99900 | 10:29 |
| *** kashyapk has quit IRC | 10:30 | |
| *** haomai___ has joined #openstack-swift | 10:32 | |
| openstackgerrit | Christian Schwede proposed a change to openstack/swift: Fix object auditor recon and logging https://review.openstack.org/99900 | 10:33 |
| *** nosnos has joined #openstack-swift | 10:34 | |
| *** NM has joined #openstack-swift | 10:34 | |
| *** mkollaro has quit IRC | 10:35 | |
| *** haomaiwa_ has quit IRC | 10:35 | |
| itarchitectkev | thanks ahale about Barbican. Probably enough for now to stave off some of the customer questions. | 10:36 |
| *** kashyapk has joined #openstack-swift | 10:38 | |
| *** nosnos has quit IRC | 10:39 | |
| *** haomaiwang has joined #openstack-swift | 10:40 | |
| *** haomaiwang has quit IRC | 10:41 | |
| *** haomai___ has quit IRC | 10:41 | |
| *** haomaiwa_ has joined #openstack-swift | 10:42 | |
| *** NM has quit IRC | 10:43 | |
| *** haomai___ has joined #openstack-swift | 10:46 | |
| *** haomaiwa_ has quit IRC | 10:46 | |
| *** haomaiwang has joined #openstack-swift | 10:48 | |
| *** haomai___ has quit IRC | 10:51 | |
| *** chandan_kumar has joined #openstack-swift | 10:51 | |
| *** mkollaro has joined #openstack-swift | 10:51 | |
| *** NM has joined #openstack-swift | 10:51 | |
| *** kashyapk has quit IRC | 10:52 | |
| *** chandankumar has quit IRC | 10:52 | |
| *** kashyapk has joined #openstack-swift | 10:52 | |
| *** NM has quit IRC | 10:54 | |
| *** chandan_kumar has quit IRC | 10:56 | |
| *** chandan_kumar has joined #openstack-swift | 10:56 | |
| *** kashyapk has quit IRC | 10:57 | |
| *** jamiehan_ has quit IRC | 11:00 | |
| *** kashyapk has joined #openstack-swift | 11:00 | |
| *** NM has joined #openstack-swift | 11:00 | |
| *** jamiehannaford has joined #openstack-swift | 11:00 | |
| *** NM has quit IRC | 11:07 | |
| *** kashyapk has quit IRC | 11:10 | |
| *** kashyapk has joined #openstack-swift | 11:10 | |
| *** kashyapk has quit IRC | 11:15 | |
| *** zul has quit IRC | 11:36 | |
| *** kashyapk has joined #openstack-swift | 11:41 | |
| *** foexle has joined #openstack-swift | 11:53 | |
| *** mmcardle has quit IRC | 11:56 | |
| *** ppai has quit IRC | 12:04 | |
| *** kashyapk has quit IRC | 12:06 | |
| *** jamiehannaford has quit IRC | 12:15 | |
| openstackgerrit | Eamonn O'Toole proposed a change to openstack/swift: Parallel object auditor https://review.openstack.org/59778 | 12:17 |
| *** jamiehannaford has joined #openstack-swift | 12:23 | |
| *** haomaiwang has quit IRC | 12:27 | |
| *** haomaiwang has joined #openstack-swift | 12:27 | |
| portante | clayg: I did not catch the 0, the probe tests did! :) | 12:29 |
| *** foexle has quit IRC | 12:32 | |
| *** haomaiw__ has joined #openstack-swift | 12:35 | |
| *** NM has joined #openstack-swift | 12:36 | |
| *** haomaiwang has quit IRC | 12:38 | |
| *** mmcardle has joined #openstack-swift | 12:42 | |
| *** kashyapk has joined #openstack-swift | 12:55 | |
| *** kashyapk has quit IRC | 12:56 | |
| *** kashyapk has joined #openstack-swift | 12:57 | |
| *** itarchitectkev_ has joined #openstack-swift | 13:15 | |
| *** itarchitectkev has quit IRC | 13:18 | |
| *** kashyapk has quit IRC | 13:21 | |
| *** foexle has joined #openstack-swift | 13:23 | |
| acoles | clayg: yeah, i thought some more and agree a relative offset from original timestamp should be sufficient for the content-type updates | 13:28 |
| acoles | clayg: so, a 64 base isn't too crazy to code but harder to debug ;) | 13:30 |
| *** zul has joined #openstack-swift | 13:32 | |
| acoles | clayg: re. 12 digit hex deca-microseconds (89 year delta) - if we add 0xa000000000000 to every offset and ditch the _ separator (split on a-f) then we get another digit for free | 13:34 |
| acoles | clayg: e.g. 1234567890.01234af9f9f9f9f9f9 . that gets us to 6 x 89 years | 13:37 |
| *** dmsimard_away is now known as dmsimard | 13:49 | |
| *** psharma has quit IRC | 13:51 | |
| *** blinky_ghost has joined #openstack-swift | 13:51 | |
| *** kashyapk has joined #openstack-swift | 14:01 | |
| openstackgerrit | Kota Tsuyuzaki proposed a change to openstack/swift: Efficient Replication for Distributed Regions https://review.openstack.org/99824 | 14:04 |
| *** lpabon has joined #openstack-swift | 14:09 | |
| *** kashyapk has quit IRC | 14:11 | |
| blinky_ghost | hi all, is it possible to use swift as mountpoint in my linux client box? | 14:14 |
| *** kevinc_ has joined #openstack-swift | 14:14 | |
| *** tdasilva has joined #openstack-swift | 14:26 | |
| *** IRTermite has joined #openstack-swift | 14:27 | |
| *** kevinc_ has quit IRC | 14:27 | |
| *** zhiyan_ is now known as zhiyan | 14:46 | |
| *** kashyapk has joined #openstack-swift | 14:53 | |
| lpabon | blinky_ghost, :-) Check out swiftonfile (previously known as Gluster-swift) | 14:56 |
| creiht | lpabon: I don't think that is exactly what he is looking for | 14:57 |
| *** Longgeek_ has quit IRC | 14:57 | |
| creiht | blinky_ghost: there is cloudfuse but it has some bugs, and I think there is a pyfuse based one as well | 14:57 |
| creiht | could look into those | 14:57 |
| *** itarchitectkev_ has quit IRC | 14:58 | |
| creiht | clayg: I thought you said you were going to change the headers back to strings? :) | 15:00 |
| *** itarchitectkev_ has joined #openstack-swift | 15:04 | |
| *** zhiyan is now known as zhiyan_ | 15:09 | |
| *** zul has quit IRC | 15:12 | |
| *** zul has joined #openstack-swift | 15:12 | |
| blinky_ghost | creiht: thanks :) | 15:14 |
| *** zul has quit IRC | 15:14 | |
| *** zul has joined #openstack-swift | 15:15 | |
| creiht | blinky_ghost: if you don't mind me asking, what were you planning on using it for? | 15:15 |
| blinky_ghost | creith: I have some remote endpoints (linux boxes) on a WAN and want to them to archive data on my datacenter storage. So I was thinking to setup Swift and use a mountpoint on the linux boxes, to put/delete/get data. | 15:19 |
| *** zul has quit IRC | 15:25 | |
| *** annegentle has quit IRC | 15:29 | |
| *** foexle has quit IRC | 15:33 | |
| *** kashyapk has quit IRC | 15:47 | |
| *** kashyapk has joined #openstack-swift | 15:48 | |
| *** kashyapk has quit IRC | 15:52 | |
| *** zul has joined #openstack-swift | 15:53 | |
| *** gyee has joined #openstack-swift | 15:58 | |
| openstackgerrit | Donagh McCabe proposed a change to openstack/swift: Add swift-checker utility for use by monitoring tools https://review.openstack.org/99961 | 16:02 |
| *** mwstorer has joined #openstack-swift | 16:04 | |
| *** zul has quit IRC | 16:14 | |
| *** NM has quit IRC | 16:24 | |
| clayg | creiht: you know that is one haven't gotten around to yet | 16:39 |
| creiht | ahh | 16:40 |
| clayg | portante: i knew there was a reason i wrote those, i got it all fixed up too, with some unittests and such - just testing all of the rebase - i did a bunch of stuff yesterday | 16:40 |
| clayg | acoles: hehehhehehe | 16:41 |
| peluse_ | clayg: BTW, once I fixed my config probe is passing for me with one ring 3x and one 2x. is there a specific test you thought would/should fail w/2x? | 16:44 |
| clayg | peluse_: i sorta thought they all would | 16:44 |
| clayg | peluse_: never got around to esting it for myself, torgomatic said his second ring was a cp of object.ring.gz, mine are all built with 3 | 16:45 |
| *** zul has joined #openstack-swift | 16:47 | |
| peluse_ | clayg: yeah, that works fine. Ran 3 times in a row last night with one 3x and one 2x and just ran again (VM just sat there overnight) and now one test fails. Sheesh | 16:47 |
| *** kashyapk has joined #openstack-swift | 16:47 | |
| clayg | heheheh | 16:47 |
| portante | clayg: great! | 16:47 |
| peluse_ | test_object_expirer_split_brain seems to be intermittent for me... will look into it a bit more and let ya know what I find | 16:50 |
| clayg | portante: i'm at a real loss on what to do with some of your comments bro? yesterday I ended up doing a lot of hand-wringing trying to figure out what to do about some of your concerns around docs and tests. | 16:50 |
| clayg | peluse_: that's the one that portante discovered the delete_at tmp dir bug on - fixed in the next rebase (coming up shortly) | 16:50 |
| *** kashyapk has quit IRC | 16:51 | |
| clayg | portante: e.g. "Are there probe or functional tests that show that all "musts" are enforced properly?" re: policy parsing | 16:51 |
| *** NM has joined #openstack-swift | 16:51 | |
| peluse_ | ahhh, cool | 16:52 |
| clayg | portante: i guess the answer is "no", although the unittests that write a file on disk and then hand it to the function that reads is and watches it blow up are about as close as you want to get to "functional" IMHO | 16:52 |
| clayg | portante: but those tests are right there in the same diff that you commented on, so I'm guessing you saw them and deamed them insufficient, and you're asking if better tests come later and .... i don't know what a better test would look like? a probetests that checks config parsing? a functional tests that can... i don't even know exactly. | 16:54 |
| clayg | portante: so I ponder and worry that I'm missing something obvious - and don't know how to respond | 16:54 |
| * portante is catching up, sec ... | 16:54 | |
| *** shri has joined #openstack-swift | 16:55 | |
| portante | clayg: I think I see your concerns, I don't think I have explained myself well enough | 16:55 |
| portante | Most of what I am concerned about is from an end user, and not a developer | 16:56 |
| peluse_ | portante: have an example if you guys don't mind be jumping in here? | 16:56 |
| portante | so as an end user, while reading the docs I am told that steps 1 - 5 will tell you if a swift cluster is setup and running correctly | 16:56 |
| clayg | portante: yeah that's great! I mean if you're using storage polices and you see something that's not implemented in a client friendly way we gotta suss it up! | 16:57 |
| clayg | portante: oh this back on the saio docs - those aren't for "end users" by any strech? | 16:57 |
| portante | so what I'd like to see, and don't want to gate anything by this, is a set of functional/probe tests that correspond to the assertiions we are documenting to users | 16:57 |
| peluse_ | portante: which specific assertions? | 16:58 |
| portante | all the steps that say: "curl ..." and you should get this, "curl ..." that and you get that | 16:58 |
| portante | so I pointed a bunch out in the various guides I was reading | 16:59 |
| peluse_ | portante: ahhhh. I can add those post merge. | 16:59 |
| clayg | portante: we have a QA Engineer, shes a real baddass, and she has some "doc driven" tests that run against our product - I guess it'd be pretty great if swift could have something similar | 16:59 |
| portante | yes, agreed | 16:59 |
| portante | peluse_: great | 16:59 |
| portante | clayg, peluse_: while I am pointing them out, I am not using a -1 to indicate we need them, just that I wanted to record them so that we can work to make that happen later. this kind of thing seems worth to do in general, not just for sp | 17:00 |
| portante | need them for sp, such that sp merge would be gated by them | 17:01 |
| *** fbo is now known as fbo_away | 17:01 | |
| *** kashyapk has joined #openstack-swift | 17:02 | |
| *** kashyapk has quit IRC | 17:07 | |
| clayg | creiht: is gholt not hanging out in here on purpose? | 17:18 |
| creiht | clayg: long story | 17:18 |
| creiht | but when he's here, he's at the computer :) | 17:19 |
| clayg | I think one of his comments says "yeah except (Exception, Timeout) will be fine" - but I'm not sure but he didn't exactly say that, it almost sounded like "I don't really care but other people care so i'm just saying" | 17:19 |
| creiht | lol | 17:19 |
| creiht | which review? | 17:19 |
| creiht | I think he mumbled something about that yesterday | 17:19 |
| creiht | something about bare excepts | 17:20 |
| clayg | https://review.openstack.org/#/c/96044/ - he was responding to me asking to confirm how i should phrase my pokemon excepts | 17:20 |
| clayg | creiht: well in a couple places someone managed to let some really silly bare excepts sneak in - there's a lot of new code in cli/info.py that is just catching WAY too much - so those I just cleaned up | 17:20 |
| creiht | yeah I think his basic stance is (which I share) | 17:21 |
| creiht | if we are doing a basic try: except Exception: block | 17:21 |
| clayg | creiht: oh oh oh get it out quick my last tests just passed I'm about to push push push! | 17:21 |
| creiht | it is usually a good idea to catch timeout there as well | 17:22 |
| creiht | as if you are catching except Exception, you are trying to catch anything bad that might happen | 17:22 |
| clayg | creiht: well yeah, the db queries all raise timeout | 17:22 |
| creiht | and while there may not be code that throws timeouts right now, that doesn't mean that someone will not add them down the road | 17:23 |
| clayg | creiht: *exactly* - it's almost like we just want to say... oh... i don't know... if on there was something that would just like catch *everything*, it's weird that python doesn't have that for just this situation... | 17:23 |
| creiht | ahh | 17:24 |
| creiht | yeah | 17:24 |
| clayg | creiht: i'm just joshin' | 17:24 |
| creiht | I'm looking at the code now | 17:24 |
| creiht | I remember a while back having bare excepts causing some wacky issues | 17:24 |
| clayg | creiht: I'm way way way fine with (Exception, Timeout) that is close enough to everything for me - i've just grown sour on it over after we had those bugs that "fixed" bare excepts | 17:25 |
| creiht | but can't remember exactly what | 17:25 |
| creiht | heh | 17:25 |
| clayg | creiht: oh... that is new | 17:25 |
| clayg | creiht: ok - so but except (Exception, Timeout) is what we're looking for there? | 17:25 |
| creiht | yeah I think that is reasonable | 17:25 |
| clayg | totally down - thanks! | 17:25 |
| creiht | cool | 17:25 |
| clayg | oh shit... here it comes! | 17:26 |
| clayg | oh... nevermind - there was a commit message I wanted to clean up - and fix authorship - 5 more mins :P | 17:27 |
| creiht | haha | 17:27 |
| creiht | I think the thing that bit us a while back, is that we had some bare exceptions hiding some really bad bugs | 17:27 |
| creiht | so we got jaded, and got very anti- bare except :) | 17:28 |
| clayg | creiht: eheheh, were they not logging the exceptions? cause that's a rule i made up as a good idea when you catch everything | 17:28 |
| creiht | clayg: yeah | 17:28 |
| creiht | things like | 17:29 |
| creiht | except: pass | 17:29 |
| creiht | :) | 17:29 |
| creiht | clayg: and I'm on board with it being fairly pragmatic that if you are going to do something like that make sure you log something | 17:29 |
| creiht | but then others might see that code and think, oh I can just put a bare except here | 17:30 |
| creiht | so it is just easier to be consistently explicit | 17:30 |
| creiht | </soapbox> | 17:31 |
| creiht | :) | 17:31 |
| *** mmcardle has quit IRC | 17:32 | |
| *** GoJkoRs has joined #openstack-swift | 17:37 | |
| GoJkoRs | Hey guys. I have a problem with Swift in horizon. Can someone point me to an guide about the integration of the two? I click on objects and it says "Error: Unable to retrieve container list." | 17:38 |
| GoJkoRs | Swift at the CLI works just fine. | 17:38 |
| *** zul has quit IRC | 17:43 | |
| *** zul has joined #openstack-swift | 17:43 | |
| *** gyee has quit IRC | 17:45 | |
| *** chuck__ has joined #openstack-swift | 17:47 | |
| *** zul has quit IRC | 17:47 | |
| *** NM has quit IRC | 18:00 | |
| clayg | ok, I think i'll really do it this time | 18:04 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Fixes probe tests with non-zero default storage policy https://review.openstack.org/96353 | 18:04 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add two vector timestamps https://review.openstack.org/99315 | 18:04 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Support https://review.openstack.org/96027 | 18:04 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Documentation https://review.openstack.org/96026 | 18:04 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy support to Containers https://review.openstack.org/96029 | 18:04 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Update FakeRing and FakeLogger https://review.openstack.org/96028 | 18:04 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy support to Object Server https://review.openstack.org/96030 | 18:04 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Fix object-expirer for missing objects https://review.openstack.org/98511 | 18:04 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add functional tests for Storage Policy https://review.openstack.org/96042 | 18:04 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Update bin scripts to be storage policy aware https://review.openstack.org/96043 | 18:04 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Merge container storage_policy_index https://review.openstack.org/96040 | 18:05 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Support to Accounts https://review.openstack.org/96041 | 18:05 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Support to list_endpoints https://review.openstack.org/96046 | 18:05 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Support to Container Sync https://review.openstack.org/96047 | 18:05 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Enqueue misplaced objects during container replication https://review.openstack.org/96044 | 18:05 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add LRUCache to common.utils https://review.openstack.org/96045 | 18:05 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Support to ssync https://review.openstack.org/96034 | 18:05 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Put X-Backend-Timestamp in object 404 responses https://review.openstack.org/96035 | 18:05 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Support to the Auditor https://review.openstack.org/96032 | 18:05 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add storage policy support for the Replicator https://review.openstack.org/96033 | 18:05 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add container-reconciler daemon https://review.openstack.org/96038 | 18:05 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add reconciler probetest outline https://review.openstack.org/96039 | 18:05 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Extend interface on InternalClient https://review.openstack.org/96036 | 18:05 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Extend direct_client https://review.openstack.org/96037 | 18:05 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy support to the Account Reaper https://review.openstack.org/96049 | 18:05 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Support to Recon Middleware https://review.openstack.org/96048 | 18:05 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy support to Object Updates https://review.openstack.org/99979 | 18:05 |
| *** chuck__ is now known as zul | 18:09 | |
| clayg | acoles: so I think the split on 'a-f' will end up sorta janky compared to an explicit seperater | 18:09 |
| *** zul has quit IRC | 18:09 | |
| *** zul has joined #openstack-swift | 18:09 | |
| *** kashyapk has joined #openstack-swift | 18:09 | |
| clayg | GoJkoRs: you may have to ask the Horizon guys - don't think many of use it. | 18:10 |
| clayg | acoles: I also think that you'll proably have to code up the overflow code (rollup the data file ts a deca-micro and buy yourself another few years of time delta) | 18:12 |
| GoJkoRs | Ok. Thank you | 18:12 |
| *** gyee has joined #openstack-swift | 18:13 | |
| *** blinky_ghost has quit IRC | 18:16 | |
| clayg | acoles: yeah I think the 12 digit fixed is the way to go, 8's just no enough to encode a reasonable number of deca-microseconds - and what's 4 bytes amoung internal use-cases | 18:17 |
| *** mkollaro has quit IRC | 18:22 | |
| clayg | k, i'm offline for a bit | 18:22 |
| *** zaitcev has joined #openstack-swift | 18:23 | |
| *** ChanServ sets mode: +v zaitcev | 18:23 | |
| *** kashyapk has quit IRC | 18:32 | |
| *** uksysadmin has joined #openstack-swift | 18:33 | |
| *** kashyapk has joined #openstack-swift | 18:33 | |
| creiht | Just saw this on the eventlet mailing list, they are calling people to test the .15 pre-release | 18:34 |
| creiht | https://lists.secondlife.com/pipermail/eventletdev/2014-June/001164.html | 18:35 |
| *** itarchitectkev_ has quit IRC | 18:35 | |
| creiht | which looks like includes a py3 branch | 18:35 |
| *** kashyapk has quit IRC | 18:37 | |
| *** uksysadmin has quit IRC | 18:38 | |
| torgomatic | huh, looks like we might actually have to start accepting all those py3 patches | 18:38 |
| *** uksysadmin has joined #openstack-swift | 18:39 | |
| creiht | hehe | 18:43 |
| *** chuck__ has joined #openstack-swift | 18:49 | |
| *** nthacker__ has joined #openstack-swift | 18:50 | |
| *** NM has joined #openstack-swift | 18:50 | |
| *** zul has quit IRC | 18:51 | |
| *** GoJkoRs has left #openstack-swift | 18:52 | |
| *** nthacker_ has quit IRC | 18:53 | |
| portante | creiht: great news | 18:56 |
| *** lpabon has quit IRC | 18:56 | |
| portante | Alex_Gaynor: curious, where are we at with PyPy and Swift? | 18:57 |
| portante | I have been using pypy for major speedups in my python code outside of swift and am liking how well it has worked so far | 18:57 |
| *** lpabon has joined #openstack-swift | 18:58 | |
| *** tdasilva has quit IRC | 18:58 | |
| *** tdasilva has joined #openstack-swift | 19:01 | |
| portante | clayg: are you going to try to get gerrit ID 100000 for the final storage policy set? :) | 19:02 |
| peluse_ | wonder if there's a prize for that... | 19:03 |
| creiht | haha | 19:04 |
| portante | probably already taken by now | 19:04 |
| peluse_ | there's goes the free toaster :( | 19:06 |
| Alex_Gaynor | portante: it works. I think it's not gatinf for side silly reason, will find our next week | 19:06 |
| *** bsdkurt has quit IRC | 19:14 | |
| *** gholt has joined #openstack-swift | 19:16 | |
| *** ChanServ sets mode: +v gholt | 19:16 | |
| gholt | ohai! | 19:17 |
| *** tdasilva has quit IRC | 19:18 | |
| portante | Alex_Gaynor: thanks | 19:22 |
| *** chuck__ has quit IRC | 19:26 | |
| *** uksysadmin has quit IRC | 19:26 | |
| *** jamiehan_ has joined #openstack-swift | 19:29 | |
| *** tdasilva has joined #openstack-swift | 19:30 | |
| *** jamiehannaford has quit IRC | 19:31 | |
| *** torgomatic has quit IRC | 19:47 | |
| *** torgomatic has joined #openstack-swift | 19:54 | |
| *** ChanServ sets mode: +v torgomatic | 19:54 | |
| *** otoolee has quit IRC | 19:55 | |
| *** DisneyRicky has quit IRC | 19:57 | |
| clayg | why isn't gerrit testing our stuff? | 19:59 |
| *** zul has joined #openstack-swift | 19:59 | |
| *** shri has quit IRC | 20:03 | |
| *** shri has joined #openstack-swift | 20:03 | |
| *** otoolee has joined #openstack-swift | 20:10 | |
| torgomatic | clayg: with the latest version of 96026 (#11), everything seems to be working for me FYI | 20:11 |
| torgomatic | dunno about gerrit; maybe it's backed up? | 20:11 |
| torgomatic | clayg: looks like it's testing to me; I'm getting emails containing "FAILURE" and "tempest-dsvm-blahblah" in close proximity | 20:15 |
| torgomatic | (notice "testing", not "working") | 20:15 |
| *** lpabon has quit IRC | 20:22 | |
| *** DisneyRicky has joined #openstack-swift | 20:24 | |
| *** miqui has quit IRC | 20:25 | |
| *** shri1 has joined #openstack-swift | 20:36 | |
| *** shri has quit IRC | 20:37 | |
| gholt | Maybe "consuming electricity" is better. | 20:38 |
| *** jamiehan_ has quit IRC | 20:39 | |
| serverascode | Hi, does anyone know if there any negative repercussions to setting object_post_as_copy = false in the proxy server configuration? | 21:02 |
| *** gholt has quit IRC | 21:05 | |
| *** tdasilva has left #openstack-swift | 21:06 | |
| creiht | serverascode: the main thing is that you can't change the content-type by a post | 21:07 |
| creiht | the client would have to change it by COPY instead | 21:07 |
| serverascode | ok thanks :) | 21:08 |
| creiht | np | 21:08 |
| creiht | container sync will not be able to sync POST updates | 21:08 |
| creiht | gif you happen to use that | 21:09 |
| serverascode | right, yeah we don't use container sync | 21:09 |
| creiht | cool | 21:09 |
| *** jamiehannaford has joined #openstack-swift | 21:15 | |
| *** haomaiw__ has quit IRC | 21:21 | |
| anticw | creiht: if POST does a copy it will | 21:31 |
| * anticw didn't read back more than 3 lines :) | 21:31 | |
| creiht | haha | 21:34 |
| anticw | i was thinking later on we can make it update the container-servers as well for POST | 21:35 |
| anticw | once the basics are working well | 21:35 |
| creiht | anticw: we talked about the possibility of doing that at the hackathon | 21:35 |
| anticw | gholt isn't here, so i think we should expect that of him | 21:35 |
| *** kashyapk has joined #openstack-swift | 21:35 | |
| creiht | so someone may be attempting something like that in the nearish future | 21:36 |
| *** kashyapk has quit IRC | 21:40 | |
| *** fbo_away is now known as fbo | 21:48 | |
| *** NM has quit IRC | 21:55 | |
| *** acoles has quit IRC | 22:03 | |
| *** acoles has joined #openstack-swift | 22:09 | |
| *** ChanServ sets mode: +v acoles | 22:09 | |
| *** jamiehannaford has quit IRC | 22:19 | |
| *** j_king has quit IRC | 22:19 | |
| *** kashyapk has joined #openstack-swift | 22:30 | |
| *** kashyapk has quit IRC | 22:35 | |
| *** dmsimard is now known as dmsimard_away | 22:51 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/python-swiftclient: Updated from global requirements https://review.openstack.org/89250 | 22:57 |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/swift: Updated from global requirements https://review.openstack.org/88736 | 22:57 |
| *** fbo is now known as fbo_away | 23:07 | |
| *** NM has joined #openstack-swift | 23:21 | |
| *** kashyapk has joined #openstack-swift | 23:33 | |
| *** kashyapk has quit IRC | 23:38 | |
| *** mwstorer has quit IRC | 23:46 | |
| *** NM has quit IRC | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!