mattoliverau | clayg: all you need to do is run a newer version of the kernel. what version are youn running? | 00:02 |
---|---|---|
clayg | mattoliverau: idk? | 00:04 |
clayg | mattoliverau: do you normally upgrade the kernal on your machines? | 00:04 |
clayg | Linux saio 3.13.0-55-generic #92-Ubuntu SMP Sun Jun 14 18:32:20 UTC 2015 x86_64 x86_64 x86_64 GNU/Linux | 00:05 |
mattoliverau | clayg: well anything above 3.15. My desktop is 4.7 | 00:05 |
clayg | but like didn't the commit message say >3.11? I wonder why my linkat tests are skipping then - oh | 00:05 |
clayg | mattoliverau: cool so xenial should do the trick - i'm all over it - thanks! | 00:06 |
mattoliverau | nps, yeah xenial that thing up :) | 00:06 |
clayg | timburke: found it! https://gist.github.com/clayg/5bc36a69b4f22de7fdb688d73af1de12 | 00:13 |
clayg | timburke: i don't *remember* writing this before? https://gist.github.com/clayg/d8e2ed3e9ad55cd287da917be5ff0efe | 00:13 |
clayg | timburke: oh bummer, i think it's just a test cleanlyness thing - probably the explody iter isn't calling close on it's sub iterators when gets close called on itself? | 00:25 |
*** tqtran has quit IRC | 00:27 | |
*** cdelatte has quit IRC | 00:29 | |
openstackgerrit | Tuan Luong-Anh proposed openstack/swift: Fix using filter() to meet python2,3 https://review.openstack.org/368649 | 00:32 |
clayg | booo - pretty sure it's just the test (not a real bug connection leak) :'( | 00:33 |
openstackgerrit | Clay Gerrard proposed openstack/swift: Fix ChunkWriteError when running unittests https://review.openstack.org/374458 | 00:34 |
openstackgerrit | Darrell Bishop proposed openstack/swift: Add support for PROXY protocol v1 (only) https://review.openstack.org/373563 | 00:47 |
*** klrmn has quit IRC | 01:26 | |
charz | morning | 01:26 |
*** bkopilov has quit IRC | 01:26 | |
zhengyin | morning | 01:26 |
zhengyin | charz: morning | 01:27 |
charz | zhengyin: o/ | 01:27 |
zhengyin | mattoliverau,kota_:morning | 01:28 |
mattoliverau | charz, zhengyin morning | 01:28 |
charz | mattoliverau: \o | 01:28 |
*** bkopilov has joined #openstack-swift | 01:32 | |
*** dmorita has quit IRC | 02:10 | |
*** StraubTW has quit IRC | 02:17 | |
*** _JZ_ has quit IRC | 02:18 | |
*** klrmn has joined #openstack-swift | 02:33 | |
*** wuhg has joined #openstack-swift | 02:39 | |
*** david-lyle has quit IRC | 03:03 | |
*** _JZ_ has joined #openstack-swift | 03:17 | |
*** _JZ_ has quit IRC | 03:49 | |
*** dmorita has joined #openstack-swift | 04:10 | |
*** tristanC has quit IRC | 04:11 | |
*** dmorita has quit IRC | 04:15 | |
*** StraubTW has joined #openstack-swift | 04:17 | |
*** ppai has joined #openstack-swift | 04:22 | |
*** StraubTW has quit IRC | 04:23 | |
*** tuanla has joined #openstack-swift | 04:42 | |
*** psachin has joined #openstack-swift | 04:44 | |
*** klrmn has quit IRC | 04:51 | |
*** klrmn has joined #openstack-swift | 05:29 | |
*** SkyRocknRoll has joined #openstack-swift | 05:31 | |
*** klrmn has quit IRC | 05:36 | |
*** Jeffrey4l has quit IRC | 05:57 | |
*** McMurlock1 has joined #openstack-swift | 05:59 | |
*** rcernin has joined #openstack-swift | 06:07 | |
*** Jeffrey4l has joined #openstack-swift | 06:10 | |
*** hseipp has joined #openstack-swift | 06:40 | |
*** oshritf has joined #openstack-swift | 07:08 | |
*** oshritf_ has joined #openstack-swift | 07:08 | |
*** asettle has joined #openstack-swift | 07:09 | |
*** ppai has quit IRC | 07:10 | |
*** rledisez has joined #openstack-swift | 07:12 | |
*** asettle has quit IRC | 07:14 | |
*** tuanla_ has joined #openstack-swift | 07:30 | |
*** tuanla has quit IRC | 07:31 | |
*** geaaru has joined #openstack-swift | 07:38 | |
*** abhitechie has joined #openstack-swift | 07:42 | |
*** acoles_ is now known as acoles | 07:46 | |
acoles | good morning | 07:50 |
mahatic | good morning acoles | 07:51 |
acoles | mahatic: I see clayg has a +2 on this https://review.openstack.org/346865, I'll take another look today and maybe we can get it merged without you needing another patchset | 07:52 |
patchbot | patch 346865 - swift - Delete old tombstones | 07:52 |
mahatic | acoles: sure, I'll do a follow-up if needed. thanks | 07:52 |
acoles | mahatic: just thinking you might want to hold off any more changes for the minor things while you have a +2 there :) | 07:52 |
mahatic | acoles: yup :) we're at the end of the week already | 07:53 |
acoles | yeah minor stuff can be cleaned up after but it would be good to get this bug fix in the release | 07:53 |
mahatic | agree | 07:54 |
*** abhinavtechie has joined #openstack-swift | 07:56 | |
*** abhitechie has quit IRC | 08:00 | |
*** oshritf_ has quit IRC | 08:08 | |
*** jamielennox is now known as jamielennox|away | 08:09 | |
*** oshritf_ has joined #openstack-swift | 08:10 | |
*** dmorita has joined #openstack-swift | 08:11 | |
*** sams-gleb has joined #openstack-swift | 08:13 | |
*** dmorita has quit IRC | 08:15 | |
*** StraubTW has joined #openstack-swift | 08:21 | |
*** tuanla has joined #openstack-swift | 08:23 | |
*** jordanP has joined #openstack-swift | 08:23 | |
*** tuanla_ has quit IRC | 08:25 | |
*** StraubTW has quit IRC | 08:26 | |
*** abhitechie has joined #openstack-swift | 08:27 | |
*** abhinavtechie has quit IRC | 08:29 | |
*** asettle has joined #openstack-swift | 08:40 | |
*** abhitechie has quit IRC | 08:59 | |
*** abhitechie has joined #openstack-swift | 08:59 | |
*** jamielennox|away is now known as jamielennox | 09:18 | |
*** ig0r_ has joined #openstack-swift | 09:25 | |
*** asettle has quit IRC | 09:26 | |
*** asettle has joined #openstack-swift | 09:27 | |
*** asettle has quit IRC | 09:28 | |
*** asettle has joined #openstack-swift | 09:28 | |
*** mvk has quit IRC | 09:35 | |
*** nikivi has joined #openstack-swift | 09:39 | |
*** amoralej has joined #openstack-swift | 09:40 | |
*** mvk has joined #openstack-swift | 10:06 | |
*** asettle has quit IRC | 10:07 | |
*** asettle has joined #openstack-swift | 10:13 | |
*** asettle has quit IRC | 10:26 | |
*** asettle has joined #openstack-swift | 10:27 | |
*** gangadhar has joined #openstack-swift | 10:35 | |
*** abhitechie has quit IRC | 10:41 | |
*** abhitechie has joined #openstack-swift | 10:43 | |
*** abhinavtechie has joined #openstack-swift | 10:46 | |
*** asettle has quit IRC | 10:48 | |
*** asettle has joined #openstack-swift | 10:48 | |
*** asettle has quit IRC | 10:48 | |
*** asettle has joined #openstack-swift | 10:49 | |
*** abhitechie has quit IRC | 10:50 | |
gangadhar | Hi everyone. I want to know rest API to login to Swift Account. I tried "swift --debug auth". But it's not showing. I didn't find it in docs either. any kind of help is appreciated. | 11:01 |
gangadhar | Got it. Thanks | 11:02 |
*** hoonetorg has joined #openstack-swift | 11:06 | |
openstackgerrit | zheng yin proposed openstack/swift: Use more specific asserts in test/unit/account tests https://review.openstack.org/342766 | 11:08 |
*** nikivi has quit IRC | 11:10 | |
*** nikivi has joined #openstack-swift | 11:14 | |
*** abhinavtechie has quit IRC | 11:20 | |
*** jistr is now known as jistr|mtg | 11:20 | |
*** ppai has joined #openstack-swift | 11:26 | |
*** nikivi has quit IRC | 11:26 | |
*** sams-gleb has quit IRC | 11:29 | |
*** sams-gleb has joined #openstack-swift | 11:29 | |
*** asettle has quit IRC | 11:53 | |
*** zigo has quit IRC | 11:54 | |
*** zigo has joined #openstack-swift | 11:58 | |
*** zigo is now known as Guest83601 | 11:59 | |
*** Guest83601 has quit IRC | 12:03 | |
*** surehs has joined #openstack-swift | 12:05 | |
*** tuanla has quit IRC | 12:06 | |
surehs | hii all, I enabled swauth in proxyserver.conf while restarting swift-proxy i am getting following issue | 12:06 |
surehs | http://paste.openstack.org/show/582569/ | 12:06 |
surehs | please some one help | 12:06 |
*** porunov has joined #openstack-swift | 12:08 | |
*** amoralej is now known as amoralej|lunch | 12:10 | |
*** zigo_ has joined #openstack-swift | 12:12 | |
*** dmorita has joined #openstack-swift | 12:12 | |
*** surehs has quit IRC | 12:16 | |
*** zigo_ has quit IRC | 12:17 | |
*** dmorita has quit IRC | 12:17 | |
*** zigo_ has joined #openstack-swift | 12:18 | |
*** abhitechie has joined #openstack-swift | 12:22 | |
*** StraubTW has joined #openstack-swift | 12:22 | |
*** StraubTW has quit IRC | 12:27 | |
*** asettle has joined #openstack-swift | 12:31 | |
*** jistr|mtg is now known as jistr | 12:36 | |
*** SkyRocknRoll_ has joined #openstack-swift | 12:37 | |
*** SkyRocknRoll has quit IRC | 12:38 | |
*** nikivi has joined #openstack-swift | 12:43 | |
*** zigo_ has quit IRC | 12:48 | |
*** zigo_ has joined #openstack-swift | 12:51 | |
acoles | has surehs gone? looks like you may need to update eventlet to >=0.18.2 before running swift-init | 12:54 |
*** david-lyle has joined #openstack-swift | 12:57 | |
*** klamath has joined #openstack-swift | 13:02 | |
*** klamath has quit IRC | 13:03 | |
*** klamath has joined #openstack-swift | 13:03 | |
*** john has joined #openstack-swift | 13:05 | |
john | Hi | 13:06 |
*** john is now known as Guest18505 | 13:06 | |
Guest18505 | I have question regarding enabling s3 api in swift | 13:06 |
Guest18505 | Can someone help me with it? | 13:07 |
Guest18505 | anybody here? | 13:07 |
*** ppai has quit IRC | 13:11 | |
mahatic | Guest18505: go ahead and post your question, whoever can help, will respond | 13:13 |
*** StraubTW has joined #openstack-swift | 13:15 | |
*** Guest18505 has quit IRC | 13:30 | |
*** psachin has quit IRC | 13:36 | |
*** gangadhar has quit IRC | 13:37 | |
*** silor has joined #openstack-swift | 13:38 | |
*** asettle has quit IRC | 13:42 | |
*** silor1 has joined #openstack-swift | 13:43 | |
*** silor has quit IRC | 13:44 | |
*** silor1 is now known as silor | 13:44 | |
*** asettle has joined #openstack-swift | 13:47 | |
acoles | ask again about s3 in US daytime | 13:59 |
*** arch-nemesis has quit IRC | 14:18 | |
*** vinsh has quit IRC | 14:20 | |
*** wuhg has quit IRC | 14:21 | |
*** daemontool has joined #openstack-swift | 14:22 | |
*** ppai has joined #openstack-swift | 14:23 | |
*** amoralej|lunch is now known as amoralej | 14:33 | |
*** nikivi has quit IRC | 14:41 | |
*** nikivi has joined #openstack-swift | 14:43 | |
*** nikivi has quit IRC | 14:43 | |
*** admin6 has joined #openstack-swift | 14:45 | |
*** ppai has quit IRC | 14:48 | |
*** cdelatte has joined #openstack-swift | 14:49 | |
*** catintheroof has joined #openstack-swift | 14:50 | |
*** daemontool has quit IRC | 14:54 | |
*** daemontool has joined #openstack-swift | 14:55 | |
*** vinsh has joined #openstack-swift | 14:57 | |
*** sams-gleb has quit IRC | 15:07 | |
*** rcernin has quit IRC | 15:07 | |
*** klrmn has joined #openstack-swift | 15:08 | |
*** zigo_ is now known as zigo | 15:14 | |
acoles | clayg: can I chat to you about patch 346865 when you are awake? | 15:15 |
patchbot | https://review.openstack.org/#/c/346865/ - swift - Delete old tombstones | 15:15 |
*** SkyRocknRoll_ has quit IRC | 15:17 | |
*** Shadow_aok has joined #openstack-swift | 15:25 | |
Shadow_aok | hello | 15:25 |
Shadow_aok | is there a way to check if a swift instance is fully synchronised ? | 15:26 |
*** chsc has joined #openstack-swift | 15:26 | |
Shadow_aok | looks like swift-recon is the tool i need | 15:36 |
mahatic | acoles: thanks for your review! Did you happen to take a look at this https://review.openstack.org/#/c/374419 | 15:39 |
patchbot | patch 374419 - swift - Move documented reclaim_age option to correct loca... | 15:39 |
*** oshritf has quit IRC | 15:40 | |
*** oshritf_ has quit IRC | 15:40 | |
mahatic | acoles: I left a comment on that for you and clayg. I'm signing off for the night | 15:41 |
acoles | mahatic: I saw it but not had chance yet to review | 15:42 |
acoles | mahatic: thanks! good night | 15:42 |
mahatic | I have some paper work to do for an appointment tomm morning. I'll check back things in my afternoon tomm (not available in the evening too) | 15:43 |
mahatic | acoles: thank you! good night too | 15:43 |
*** nikivi has joined #openstack-swift | 15:43 | |
mahatic | evening* | 15:43 |
*** Shadow_aok has quit IRC | 15:44 | |
*** tongli has joined #openstack-swift | 15:47 | |
*** schizo is now known as bwall | 15:51 | |
*** takashi has joined #openstack-swift | 15:54 | |
*** nikivi has quit IRC | 15:56 | |
openstackgerrit | Alistair Coles proposed openstack/swift: Cleanup tests for auditor invalidating hashes https://review.openstack.org/374957 | 16:07 |
openstackgerrit | Darrell Bishop proposed openstack/swift: Add support for PROXY protocol v1 (only) https://review.openstack.org/373563 | 16:12 |
*** dmorita has joined #openstack-swift | 16:12 | |
*** chsc has quit IRC | 16:12 | |
notmyname | hello, world | 16:14 |
notmyname | sorry i couldn't get online yesterday. trying to catch up now | 16:14 |
*** dmorita has quit IRC | 16:17 | |
*** rcernin has joined #openstack-swift | 16:20 | |
*** rledisez has quit IRC | 16:22 | |
timburke | good morning | 16:24 |
*** _JZ_ has joined #openstack-swift | 16:25 | |
*** hseipp has quit IRC | 16:25 | |
pdardeau | jrichli: acoles: mahatic: did you guys look at GCM mode of AES when you did the encryption work? | 16:32 |
notmyname | acoles: oh, you're wrong | 16:36 |
notmyname | acoles: about the authtoken config section in our sample configs ;-) | 16:37 |
*** silor has quit IRC | 16:37 | |
*** jordanP has quit IRC | 16:39 | |
acoles | notmyname: heh. welcome back | 16:39 |
notmyname | :-) | 16:40 |
notmyname | I need to check on the status of patches for the release | 16:40 |
notmyname | haven't gone through all the gerrit emails yet | 16:40 |
notmyname | but I saw there were a few more patches that landed | 16:40 |
*** zigo has quit IRC | 16:41 | |
*** mvk has quit IRC | 16:41 | |
*** asettle has quit IRC | 16:41 | |
notmyname | for patch https://review.openstack.org/#/c/346865/, we could do that as-is, and mahatic could fix the small things in a follow up? what are your thoughts on that acoles? | 16:42 |
patchbot | patch 346865 - swift - Delete old tombstones | 16:42 |
*** asettle has joined #openstack-swift | 16:42 | |
*** dmorita has joined #openstack-swift | 16:42 | |
*** geaaru has quit IRC | 16:43 | |
*** dmorita has quit IRC | 16:44 | |
*** dmorita has joined #openstack-swift | 16:44 | |
*** zul has quit IRC | 16:44 | |
acoles | notmyname: agree re small fixes being a follwo up but I have one concern that we may want to fix first, I just posted a follow-up to address it... | 16:45 |
acoles | https://review.openstack.org/374957 | 16:45 |
patchbot | patch 374957 - swift - Cleanup tests for auditor invalidating hashes | 16:45 |
*** zul has joined #openstack-swift | 16:45 | |
notmyname | ah, great | 16:45 |
acoles | hmm the commit subject disguises that there is a fix in there | 16:45 |
*** asettle has quit IRC | 16:46 | |
acoles | notmyname: I was waiting to discuss with clayg and have him tell me I am fretting about nothing :) but it does seem unnecessary to call invalidate_hash twice | 16:46 |
notmyname | ok :-) | 16:46 |
notmyname | I'll work on the authors/changelog update today (probably on the flight) | 16:47 |
acoles | notmyname: so I would be +2 on patch 346865 if the fix in patch 374957 were squashed in, or if clayg slap me hard enough | 16:47 |
patchbot | https://review.openstack.org/#/c/346865/ - swift - Delete old tombstones | 16:47 |
patchbot | https://review.openstack.org/#/c/374957/ - swift - Cleanup tests for auditor invalidating hashes | 16:47 |
acoles | unfortunately my follow-up reworks the tests, but I kinda needed that to get a test for the fix | 16:48 |
notmyname | I'll keep an eye on those and then propose my authors/changelog patch tonight. which means that we should have a SHA for the release tomorrow am (california time) | 16:50 |
notmyname | seem reasonable? | 16:50 |
*** zigo has joined #openstack-swift | 16:51 | |
acoles | notmyname: yes | 16:51 |
*** zigo is now known as Guest18656 | 16:52 | |
notmyname | zigo: FYI, that ^. expect to see a swift release available tomorrow am pacific time | 16:52 |
notmyname | (assuming the -release team is good with that) | 16:52 |
acoles | notmyname: we've had some discussion on this one https://review.openstack.org/#/c/372573, as it now stands it does not make any change but beefs up the comments | 16:52 |
patchbot | patch 372573 - swift - Add more comment to authtoken sample options | 16:52 |
jrichli | pdardeau: not that I know of. although cca did in the beginning want crypto + auth/HMAC, but I think it was CBC + HMAC. | 16:52 |
acoles | jrichli: thanks! I was about to attempt a reply to pdardeau but you got it. IIRC the concern was that a scheme that adds authentication info to verify integrity results in changed payload size, which makes ranged GETs a headache. | 16:54 |
jrichli | pdardeau: I am reading about GCM to try to see if it also is positionally equivalent so that there is not size increase of the message | 16:55 |
notmyname | ah, timburke just pointed out https://review.openstack.org/#/c/373537/ to me | 16:55 |
patchbot | patch 373537 - swift - Use separate headers for versioned_writes' stack a... | 16:55 |
timburke | notmyname: don't worry too much; jrichli, clayg, and tdasilva are all interested | 16:55 |
timburke | and mattoliverau already +2ed, so... | 16:56 |
*** Guest18656 has quit IRC | 16:56 | |
notmyname | interested enough to review/land today? | 16:56 |
*** Jeffrey4l has quit IRC | 16:57 | |
timburke | well that's always the question, isn't it? ;-) | 16:57 |
jrichli | as for me, I have started the review, but still have halfway to go. | 16:57 |
jrichli | I should be mostly freed up to do that after lunch | 16:58 |
*** zigo_ has joined #openstack-swift | 16:59 | |
timburke | jrichli: on the having authenticated encryption, maybe the auth piece could go in object sysmeta? how big is that data, typically? | 17:00 |
*** nikivi has joined #openstack-swift | 17:06 | |
*** chsc has joined #openstack-swift | 17:07 | |
*** zigo_ has quit IRC | 17:07 | |
*** chsc has quit IRC | 17:07 | |
*** chsc has joined #openstack-swift | 17:07 | |
*** tongli_ has joined #openstack-swift | 17:07 | |
clayg | heyoh! | 17:08 |
notmyname | yo | 17:09 |
*** tongli has quit IRC | 17:10 | |
*** zigo_ has joined #openstack-swift | 17:11 | |
*** tongli_ has quit IRC | 17:12 | |
*** chsc has quit IRC | 17:12 | |
acoles | clayg: hi. so I was about to +2A patch 346865 but then noticed duplicate suffixes appearing in my hashes.invalid files | 17:13 |
patchbot | https://review.openstack.org/#/c/346865/ - swift - Delete old tombstones | 17:13 |
acoles | clayg: cos the invalidate_hash is getting called by the zero bytes files auditor process and the regular process | 17:14 |
acoles | clayg: I put a fix here https://review.openstack.org/#/c/374957/ , question is, am I fretting too much? do we land patch 346865 as it is or get rid of the duplication? | 17:15 |
patchbot | https://review.openstack.org/#/c/346865/ - swift - Delete old tombstones | 17:15 |
patchbot | patch 374957 - swift - Cleanup tests for auditor invalidating hashes | 17:15 |
*** ChubYann has joined #openstack-swift | 17:17 | |
*** chsc has joined #openstack-swift | 17:17 | |
*** chsc has joined #openstack-swift | 17:17 | |
zaitcev | Interesting. I certainly didn't realize those dupes could happen. | 17:21 |
zaitcev | BTW | 17:22 |
zaitcev | Do you guys remember that XFS people (well, Dave Chinner and Ric Wheeler) were mad about the zero byte files. They maintained that it's not possible if app issues the right flushes in the right order, and so our ZBF thing basically cleans up bugs in Swift not closing files properly. | 17:23 |
zaitcev | If we're doing everything right, maybe we can get rid of ZBF | 17:23 |
zaitcev | Is there a stat if it actually finds zero byte files in recent clusters e.g. 2.7.0) | 17:24 |
notmyname | zaitcev: is the "right" way along the lines of the O_TMPFILE patch that recently landed? | 17:24 |
zaitcev | notmyname: I need to find out. | 17:24 |
notmyname | I like trying to find out what ZBF scans find, too | 17:25 |
*** McMurlock1 has quit IRC | 17:27 | |
acoles | clayg: NM, I have to leave. tell me what you think about https://review.openstack.org/#/c/346865/ on gerrit and I can do whatever is needed my daytime tomorrow so we get it landed in time for notmyname to release | 17:29 |
patchbot | patch 346865 - swift - Delete old tombstones | 17:29 |
acoles | good night | 17:29 |
notmyname | good night | 17:29 |
*** acoles is now known as acoles_ | 17:29 | |
*** nadeem has joined #openstack-swift | 17:34 | |
notmyname | FYI https://etherpad.openstack.org/p/BCN-ops-meetup is what people are using to help schedule the ops sessions in barcelona. take a look and see what's interesting (there is one proposed topic on swift) | 17:38 |
*** amoralej is now known as amoralej|off | 17:43 | |
*** takashi has quit IRC | 17:55 | |
*** tqtran has joined #openstack-swift | 17:59 | |
*** ig0r_ has quit IRC | 18:09 | |
*** dmsimard is now known as dmsimard|afk | 18:10 | |
clayg | ... sorry acoles_ - so the cycle times of the zbf and the regular auditor are so vastly different on loaded disks I don't think it's really going to make that much of a difference (a REPLICATE verb will rehash the suffix before a duplicate entry can sneak in) | 18:18 |
clayg | but I like the guard regardless - seems like a nice improvment | 18:18 |
clayg | mahatic: I think this is telling me you're about done for the day -> http://stackalytics.com/report/users/mahati-chamarthy | 18:20 |
*** tgtanya has joined #openstack-swift | 18:20 | |
*** tgtanya_ has joined #openstack-swift | 18:21 | |
*** nikivi has quit IRC | 18:22 | |
jrichli | timburke: I'd have to do some research before I could chat on it. If you all think that is a use case people are interested, I could learn about it. | 18:23 |
timburke | jrichli: honestly, i don't really know much about authenticated encryption myself. getting through the implementation we have now only required little refreshers; i think i'd need to do some more concentrated research to update it, too | 18:31 |
*** tgtanya_ has quit IRC | 18:33 | |
zaitcev | I tend to like what mahatic did with old tombstones, but mostly because I had them observed in the wild. Possibly my cluster didn't have enough health, but I accumulated a ton of them. Like tens of thousands. | 18:37 |
openstackgerrit | Clay Gerrard proposed openstack/swift: wip fix wsgi refactor https://review.openstack.org/375056 | 18:47 |
*** tgtanya has quit IRC | 18:50 | |
clayg | zaitcev: at this point it's really more of a stats-only checker vs. a do the md5 checker | 18:53 |
clayg | zaitcev: I recently saw a bunch of corrupt data that was "filled with nulls" - basically it looked like a sparse file. | 18:53 |
*** porunov has quit IRC | 18:53 | |
clayg | either way I mostly think the quarantiner is supposed to deal with like physical spinning rust read errors? How can the file system be expected to "do the right thing" when the underlying block device coughs up bullshit? Who knows what horrors hard-driver manufacturers sneak into their drivers to get their disks to pass their read error tests? | 18:56 |
clayg | I do know that it's not rare for size of the file reported by the filesystem (by corrupt lying broken spinning rust) often != the size we wrote into the metadta (most often some kind of truncation - I'm not sure how often the stat is really *zero*; vs just "smaller") | 18:58 |
clayg | another possibility is something with the rename operation was at one time contributing to the zero-byte hypothosis (missing fsync on the newly created quarnatine subdirs?) | 19:00 |
clayg | ... but in general I think running the metadata only scan faster than the checksum scan is a huge win that's catching 80-90% of the bad objects that need to be kicked out and repaired from replication | 19:01 |
clayg | ... but the name may be more superstitious than some sort of evidence based approach | 19:01 |
clayg | https://cdn.meme.am/instances/500x/30655394.jpg maybe? | 19:03 |
zaitcev | Well yea, but the question is if a dedicated ZBF does anything that normal auditor won't. | 19:03 |
clayg | zaitcev: and i'm tell yes, it's catching most of what i'm seeing quarantined and it's doing it way faster than the checksum scan | 19:03 |
zaitcev | clayg: understood | 19:04 |
clayg | zaitcev: it doesn't *only* look for zero byte files - it checks for size from stat != size in metadata | 19:04 |
*** daemontool has quit IRC | 19:12 | |
*** daemontool has joined #openstack-swift | 19:13 | |
*** tongli has joined #openstack-swift | 19:14 | |
*** tongli_ has joined #openstack-swift | 19:17 | |
*** daemontool_ has joined #openstack-swift | 19:18 | |
*** daemontool has quit IRC | 19:19 | |
*** tongli has quit IRC | 19:19 | |
*** daemontool_ has quit IRC | 19:24 | |
timburke | clayg: on https://review.openstack.org/#/c/374957, do we know why acoles_ went with ALL instead of ZBF? looks like his original recommendation was the ZBF...? | 19:25 |
patchbot | patch 374957 - swift - Cleanup tests for auditor invalidating hashes | 19:25 |
*** daemontool_ has joined #openstack-swift | 19:25 | |
*** zigo_ is now known as zigo | 19:30 | |
notmyname | hello from 32k feet | 19:37 |
timburke | notmyname: ready to talk about clouds? | 19:38 |
pdardeau | we've got friends in high places :) | 19:38 |
notmyname | this is actually the first time I've ever bought wifi on a plane | 19:38 |
notmyname | ok, so based on what acoles_ was saying earlier, I think we can take his +1 as the second approval and land https://review.openstack.org/#/c/346865/ | 19:47 |
patchbot | patch 346865 - swift - Delete old tombstones | 19:47 |
notmyname | clayg: timburke: is that your interpretation too, or should we wait until his tomorrow? | 19:47 |
notmyname | as I saw it, acoles_ is good with patch 346865 as long as the stuff in patch 374957 lands too | 19:48 |
patchbot | https://review.openstack.org/#/c/346865/ - swift - Delete old tombstones | 19:48 |
patchbot | https://review.openstack.org/#/c/374957/ - swift - Cleanup tests for auditor invalidating hashes | 19:48 |
timburke | seemed to be what he was saying, but i hesitated from just +A'ing the other patch since i haven't looked at it | 19:49 |
timburke | i just trust that it behaves properly because acoles_ and clayg seem to say so :-) | 19:49 |
notmyname | :-) | 19:49 |
notmyname | I'm happy with doing +A if we agree that is acoles_'s intention | 19:49 |
notmyname | tdasilva: thanks for reviewing https://review.openstack.org/#/c/373537/ | 19:50 |
patchbot | patch 373537 - swift - Use separate headers for versioned_writes' stack a... | 19:50 |
tdasilva | notmyname: yw, thanks to timburke for getting it done so quickly... | 19:51 |
tdasilva | notmyname: have a safe flight! | 19:52 |
notmyname | tdasilva: I'm just sitting here. the safety is dependent on (1) the pilot and (2) nobody turning on their Note 7 ;-) | 19:52 |
tdasilva | lol...also make sure there are no kids with their hoverboards going up and down the aisles :P | 19:53 |
notmyname | funny/scary how many times the airport is announcing "due to the recall [ie exploding], do not turn a Note 7 on. do not plug it in. do not feed after midnight and do not taunt happy fun note 7" | 19:53 |
*** oshritf_ has joined #openstack-swift | 19:58 | |
*** oshritf has joined #openstack-swift | 19:58 | |
jrichli | I just watched Gremlins the other day :-) | 19:58 |
openstackgerrit | Merged openstack/swift: Fix ChunkWriteError when running unittests https://review.openstack.org/374458 | 20:00 |
notmyname | it's been a *long* time since I've seen it | 20:04 |
openstackgerrit | Merged openstack/swift: Use separate headers for versioned_writes' stack and history modes https://review.openstack.org/373537 | 20:05 |
notmyname | timburke: yay! ^ | 20:05 |
notmyname | random meta thought about IRC: does everyone in here see messages in the same order when they happen close together? what are the guarantees the ircd daemon (or freenode?) offer there? how do they have a giant distributed system and handle conversations (which are inherently ordered)? | 20:06 |
notmyname | meta meta thought: it's amazing anything works at all | 20:07 |
* notmyname ponders as he sits in a metal tube miles up in the air talking to people, in real time, all over the world | 20:08 | |
notmyname | mostly https://www.youtube.com/watch?v=q8LaT5Iiwo4 | 20:08 |
bkeller` | i don't think anything's guaranteed to be in order since the servers are distributed and send messages as soon as they see them | 20:13 |
zaitcev | IRC servers form a spanning tree, so unless the network is recovering from splits, there will be no out of order messages. | 20:14 |
notmyname | ah, interesting | 20:15 |
bkeller` | oh, i'm wrong then | 20:15 |
notmyname | onovy: I saw you say something in a different channel. did you see my comment on the swift release? the plan is to get it tagged tomorrow | 20:15 |
bkeller` | maybe i'm just too used to splits :\ | 20:15 |
onovy | notmyname: hi | 20:15 |
onovy | notmyname: cool! missed that | 20:16 |
tdasilva | notmyname: hopefully you won't sit on the tarmac for 40 min ;) | 20:16 |
onovy | i have few reviews, make they can land before? :) | 20:16 |
onovy | for example: https://review.openstack.org/#/c/240726/ and https://review.openstack.org/#/c/340250/ | 20:17 |
patchbot | patch 240726 - swift - Respect server type for --md5 check in swift-recon | 20:17 |
patchbot | patch 340250 - swift - Add a jitter for check of time synchronization. | 20:17 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/swift: Updated from global requirements https://review.openstack.org/88736 | 20:20 |
*** oshritf has quit IRC | 20:22 | |
*** oshritf_ has quit IRC | 20:22 | |
*** tongli_ has quit IRC | 20:23 | |
jrichli | notmyname: great clip | 20:24 |
*** tongli has joined #openstack-swift | 20:29 | |
clayg | louis ck is really great - my favorite is when he does like horribly offensive jokes that aren't funny - he just gets a kick out of watching the audience crindge - it's like a joke just for him | 20:30 |
clayg | timburke: i'm not sure why acoles_ picked which worker - i don't think he cares (i don't really care?) - either is fine probably - i don't really have any context to feel one is better than the other? | 20:34 |
clayg | notmyname: I think al's follow up is really part of the same change - the probably should be squashed; OTOH I think swift is correct and complete in both sha's tho - so we can merge them seperately if it's helpful to do so | 20:36 |
notmyname | clayg: ack. since the follow-up already has the gerrit bits, let's just land them both | 20:36 |
notmyname | I'll click the +A for it | 20:36 |
clayg | wfm | 20:37 |
*** cdelatte has quit IRC | 20:37 | |
*** abhitechie has quit IRC | 20:38 | |
*** asettle has joined #openstack-swift | 20:51 | |
*** airdisa has joined #openstack-swift | 20:52 | |
notmyname | anyone played with using prometheus to monitor swift? I've heard a few people ask about it, so I wanted to try it out | 20:53 |
mattoliverau | morning | 20:56 |
notmyname | hello mattoliverau | 20:57 |
mattoliverau | notmyname: hey man, how's the travels? | 20:57 |
notmyname | almost done. in a plane now | 20:57 |
mattoliverau | Cool | 20:58 |
*** dmsimard|afk is now known as dmsimard | 21:00 | |
jrichli | notmyname: were there any interesting questions about crypto after the talks? | 21:11 |
notmyname | FIPS FIPS FIPS FIPS FISMA FIPS FIPS FIPS | 21:11 |
notmyname | no, the oxygen masks didn't just deploy. those are government standards about security | 21:12 |
notmyname | jrichli: oh, one questions I didn't remember the answer to: why don't we encrypt user metadata keys? | 21:12 |
jrichli | heh. I don't personally know the answer either. That decision was in place before I started, and I just assumed it would be too difficult. Since then, I guess I have learned difficult wouldn't stop swift ;-) | 21:14 |
jrichli | I am glad I didn't have to field the questions on compliance :-) | 21:15 |
notmyname | did you have to deal with any of those in a previous life? | 21:17 |
notmyname | back when you were doing those secret things you aren't allowed to talk about | 21:18 |
jrichli | they had people that took care of those things. i just wrote code :-) | 21:18 |
notmyname | heh, ok | 21:19 |
jrichli | we had one of those questions in Austin, but it was after the session was over so it was just a one-on-one discussion basically. not so hard to answer that way, IMO | 21:20 |
notmyname | the generic answer is that it's hard to think about getting an open source project certified, but it's something we can look in to | 21:22 |
*** tongli has quit IRC | 21:23 | |
*** gyee has joined #openstack-swift | 21:27 | |
openstackgerrit | Merged openstack/swift: Delete old tombstones https://review.openstack.org/346865 | 21:29 |
openstackgerrit | Merged openstack/swift: Cleanup tests for auditor invalidating hashes https://review.openstack.org/374957 | 21:31 |
*** rcernin has quit IRC | 21:34 | |
*** zul has quit IRC | 21:36 | |
*** admin6 has quit IRC | 21:36 | |
*** bkopilov has quit IRC | 21:37 | |
*** chianingwang has quit IRC | 21:37 | |
*** ntata has quit IRC | 21:37 | |
*** amitkqed has quit IRC | 21:37 | |
*** daemontool_ has quit IRC | 21:37 | |
*** chsc has quit IRC | 21:37 | |
*** klrmn has quit IRC | 21:37 | |
*** clayg has quit IRC | 21:37 | |
*** jlvillal has quit IRC | 21:37 | |
*** onovy has quit IRC | 21:37 | |
*** DuncanT has quit IRC | 21:37 | |
*** _JZ_ has quit IRC | 21:37 | |
*** klamath has quit IRC | 21:37 | |
*** acoles_ has quit IRC | 21:37 | |
*** blair has quit IRC | 21:37 | |
*** tdasilva has quit IRC | 21:37 | |
*** charz has quit IRC | 21:37 | |
*** markstur has quit IRC | 21:37 | |
*** madorn has quit IRC | 21:37 | |
*** kozhukalov has quit IRC | 21:37 | |
*** calebb has quit IRC | 21:37 | |
*** tries_ has quit IRC | 21:37 | |
*** kragniz has quit IRC | 21:37 | |
*** fbo has quit IRC | 21:37 | |
*** etienneme has quit IRC | 21:37 | |
*** samueldmq has quit IRC | 21:37 | |
*** zhiyan has quit IRC | 21:37 | |
*** mahatic has quit IRC | 21:37 | |
*** oxinabox has quit IRC | 21:37 | |
*** zackmdavis has quit IRC | 21:37 | |
*** tqtran has quit IRC | 21:37 | |
*** torgomatic has quit IRC | 21:37 | |
*** pdardeau has quit IRC | 21:37 | |
*** clarkb has quit IRC | 21:37 | |
*** hugokuo has quit IRC | 21:37 | |
*** ajiang has quit IRC | 21:37 | |
*** dosaboy has quit IRC | 21:37 | |
*** bobby2 has quit IRC | 21:37 | |
*** chmouel has quit IRC | 21:37 | |
*** niknakpaddywak has quit IRC | 21:37 | |
*** stevemar has quit IRC | 21:37 | |
*** aj701 has quit IRC | 21:37 | |
*** ndk- has quit IRC | 21:37 | |
*** balajir has quit IRC | 21:37 | |
*** gmmaha has quit IRC | 21:37 | |
*** StraubTW has quit IRC | 21:37 | |
*** jroll has quit IRC | 21:37 | |
*** yarkot has quit IRC | 21:37 | |
*** sgundur has quit IRC | 21:37 | |
*** a1|away has quit IRC | 21:37 | |
*** tanee has quit IRC | 21:37 | |
*** jamielennox has quit IRC | 21:37 | |
*** jrichli has quit IRC | 21:37 | |
*** CrackerJackMack has quit IRC | 21:37 | |
*** tmoreira has quit IRC | 21:37 | |
*** cargonza has quit IRC | 21:37 | |
*** amit213 has quit IRC | 21:37 | |
*** nottrobin has quit IRC | 21:37 | |
*** tamizh_geek has quit IRC | 21:37 | |
*** zacksh has quit IRC | 21:37 | |
*** rsFF has quit IRC | 21:37 | |
*** _JZ_ has joined #openstack-swift | 21:38 | |
*** klamath has joined #openstack-swift | 21:38 | |
*** acoles_ has joined #openstack-swift | 21:38 | |
*** blair has joined #openstack-swift | 21:38 | |
*** tdasilva has joined #openstack-swift | 21:38 | |
*** charz has joined #openstack-swift | 21:38 | |
*** markstur has joined #openstack-swift | 21:38 | |
*** madorn has joined #openstack-swift | 21:38 | |
*** tries_ has joined #openstack-swift | 21:38 | |
*** kozhukalov has joined #openstack-swift | 21:38 | |
*** calebb has joined #openstack-swift | 21:38 | |
*** kragniz has joined #openstack-swift | 21:38 | |
*** fbo has joined #openstack-swift | 21:38 | |
*** samueldmq has joined #openstack-swift | 21:38 | |
*** oxinabox has joined #openstack-swift | 21:38 | |
*** etienneme has joined #openstack-swift | 21:38 | |
*** zhiyan has joined #openstack-swift | 21:38 | |
*** mahatic has joined #openstack-swift | 21:38 | |
*** zackmdavis has joined #openstack-swift | 21:38 | |
*** orwell.freenode.net sets mode: +v acoles_ | 21:38 | |
*** kong has quit IRC | 21:39 | |
*** ediardo has quit IRC | 21:39 | |
*** asettle is now known as 7JTABTGDJ | 21:40 | |
*** asettle has joined #openstack-swift | 21:40 | |
*** StraubTW has joined #openstack-swift | 21:40 | |
*** jroll has joined #openstack-swift | 21:40 | |
*** sgundur has joined #openstack-swift | 21:40 | |
*** yarkot has joined #openstack-swift | 21:40 | |
*** a1|away has joined #openstack-swift | 21:40 | |
*** tanee has joined #openstack-swift | 21:40 | |
*** jamielennox has joined #openstack-swift | 21:40 | |
*** jrichli has joined #openstack-swift | 21:40 | |
*** CrackerJackMack has joined #openstack-swift | 21:40 | |
*** tmoreira has joined #openstack-swift | 21:40 | |
*** amit213 has joined #openstack-swift | 21:40 | |
*** nottrobin has joined #openstack-swift | 21:40 | |
*** tamizh_geek has joined #openstack-swift | 21:40 | |
*** zacksh has joined #openstack-swift | 21:40 | |
*** rsFF has joined #openstack-swift | 21:40 | |
*** orwell.freenode.net sets mode: +v jrichli | 21:40 | |
*** nottrobin has quit IRC | 21:40 | |
*** zul has joined #openstack-swift | 21:41 | |
*** admin6 has joined #openstack-swift | 21:41 | |
*** bkopilov has joined #openstack-swift | 21:41 | |
*** chianingwang has joined #openstack-swift | 21:41 | |
*** ntata has joined #openstack-swift | 21:41 | |
*** amitkqed has joined #openstack-swift | 21:41 | |
*** serverascode has quit IRC | 21:41 | |
*** cppforlife_ has quit IRC | 21:41 | |
*** 7JTABTGDJ has quit IRC | 21:41 | |
*** AndyWojo has quit IRC | 21:41 | |
*** daemontool_ has joined #openstack-swift | 21:41 | |
*** chsc has joined #openstack-swift | 21:41 | |
*** klrmn has joined #openstack-swift | 21:41 | |
*** clayg has joined #openstack-swift | 21:41 | |
*** jlvillal has joined #openstack-swift | 21:41 | |
*** onovy has joined #openstack-swift | 21:41 | |
*** orwell.freenode.net sets mode: +v clayg | 21:41 | |
*** zhiyan has quit IRC | 21:41 | |
*** daemontool_ has quit IRC | 21:42 | |
*** tqtran has joined #openstack-swift | 21:42 | |
*** torgomatic has joined #openstack-swift | 21:42 | |
*** pdardeau has joined #openstack-swift | 21:42 | |
*** clarkb has joined #openstack-swift | 21:42 | |
*** hugokuo has joined #openstack-swift | 21:42 | |
*** ajiang has joined #openstack-swift | 21:42 | |
*** niknakpaddywak has joined #openstack-swift | 21:42 | |
*** dosaboy has joined #openstack-swift | 21:42 | |
*** bobby2 has joined #openstack-swift | 21:42 | |
*** chmouel has joined #openstack-swift | 21:42 | |
*** stevemar has joined #openstack-swift | 21:42 | |
*** aj701 has joined #openstack-swift | 21:42 | |
*** ndk- has joined #openstack-swift | 21:42 | |
*** balajir has joined #openstack-swift | 21:42 | |
*** gmmaha has joined #openstack-swift | 21:42 | |
*** orwell.freenode.net sets mode: +v torgomatic | 21:42 | |
*** amit213 has quit IRC | 21:44 | |
*** kozhukalov has quit IRC | 21:44 | |
*** asettle has quit IRC | 21:45 | |
*** ediardo has joined #openstack-swift | 21:47 | |
*** kong has joined #openstack-swift | 21:51 | |
notmyname | clayg: why'd you make me watch something by raymond hettinger. I've seen this many times alreayd, but I just can't stop watching. so good. | 21:51 |
*** nottrobin has joined #openstack-swift | 21:53 | |
*** cppforlife_ has joined #openstack-swift | 21:57 | |
*** mvk has joined #openstack-swift | 21:58 | |
*** nikivi has joined #openstack-swift | 21:58 | |
openstackgerrit | Mathias Bjoerkqvist proposed openstack/swift: WIP: Storing encryption root secret in Barbican https://review.openstack.org/364878 | 22:00 |
*** abhitechie has joined #openstack-swift | 22:03 | |
*** cargonza has joined #openstack-swift | 22:04 | |
*** nadeem has quit IRC | 22:05 | |
*** StraubTW has quit IRC | 22:07 | |
*** DuncanT has joined #openstack-swift | 22:08 | |
*** nikivi has quit IRC | 22:10 | |
*** kozhukalov has joined #openstack-swift | 22:10 | |
*** AndyWojo has joined #openstack-swift | 22:11 | |
*** amit213 has joined #openstack-swift | 22:14 | |
*** serverascode has joined #openstack-swift | 22:14 | |
*** zhiyan has joined #openstack-swift | 22:16 | |
clayg | notmyname: yeah my bad | 22:24 |
clayg | he's great | 22:24 |
*** abhitechie has quit IRC | 22:35 | |
*** chsc has quit IRC | 22:35 | |
*** airdisa has quit IRC | 22:38 | |
*** airdisa has joined #openstack-swift | 22:38 | |
*** vinsh has quit IRC | 22:41 | |
*** mvk has quit IRC | 22:52 | |
*** StraubTW has joined #openstack-swift | 22:53 | |
*** StraubTW_ has joined #openstack-swift | 22:54 | |
*** StraubTW has quit IRC | 22:58 | |
*** StraubTW_ has quit IRC | 23:01 | |
*** airdisa has quit IRC | 23:06 | |
*** mvk has joined #openstack-swift | 23:08 | |
openstackgerrit | Darrell Bishop proposed openstack/swift: Add support for PROXY protocol v1 (only) https://review.openstack.org/373563 | 23:13 |
*** kei_yama has joined #openstack-swift | 23:21 | |
*** abhitechie has joined #openstack-swift | 23:32 | |
*** hosanai has joined #openstack-swift | 23:45 | |
*** ChanServ sets mode: +v hosanai | 23:45 | |
*** ppai has joined #openstack-swift | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!