Wednesday, 2015-09-02

*** lpabon has quit IRC00:02
*** kota_ has joined #openstack-swift00:04
*** ChanServ sets mode: +v kota_00:04
kota_good morning00:04
ctennispeluse: I'm trying to merge in the various EC fixes from master to this test cluster I have, but the two patchsets https://review.openstack.org/#/c/212187/ https://review.openstack.org/#/c/213147/ seem to have conflicts.00:05
claygctennis: https://review.openstack.org/#/c/212187 is a for sure for sure00:06
ctennisok cool00:07
peluseone sec...00:07
peluseOK, I'll rebase the alt frags thing now00:08
clayghttps://review.openstack.org/#/c/213147 needs to be revisited - I'm throwing my money in with https://review.openstack.org/#/c/215276/ over back tracking to nodes for alt-frag-indexes as far as value - but that's somewhat based on my judgement of hitting lp bug #1469094 being pretty good00:09
openstackLaunchpad bug 1469094 in OpenStack Object Storage (swift) "Timeout writing .durable can cause error on GET (under failure)" [High,Confirmed] https://launchpad.net/bugs/146909400:09
claygpeluse: FWIW i tried to rebase the alt-frag get stuff and found it somewhat devestated by acoles_ and my changes to the etag_buckets proxy path00:09
peluseclayg, oh boy00:10
claygpeluse: ctennis: did you guys *see* any duplicate frags on the nodes when you were running?00:10
pelusenot that we know of but we didn't get that far in testing00:10
ctennisduplicate as in side by side in the same physical spot?00:10
claygas we understand it to get doubled frags you need a) rebalance & failures or b) reconstructor bugs00:10
claygwhere reconstructor bugs in the early days was like "pyeclib segfaults so the reconstructor isn't running for days"00:11
ctennisnever saw two frags in the same place00:11
claygctennis: yup ts0#2.data and ts0#3.data in the same hashdir00:11
peluseyeah, duplicates is probably not the right word.  mult frag indexes in same hash dir00:11
ctennisyeah, I don't know we ever got far enough to feel confident in answering that, but with what little I looked it I never saw it00:12
*** darrenc is now known as darrenc_afk00:12
claygwell anyway - I think backtracking to nodes for alt-frag indexes would be sweet - doing the bare minimum of not letting in two frag index for the same bucket would be a GREAT patch (let the parity nodes and buckets do it's thing) - but I'm honestly a little worried about the durable timeouts causing some issues for failure testing - or even normal benchmarking if you drive it hard enough00:13
peluseclayg, rebase doesn't look *that* bad... not sure I know what you mean by backtracking though00:14
claygbacktracking - reinserting nodes into the iter so you can make a second connection to the same node00:14
clayggoing - back - after the train is already down the track - backtracking :P00:14
peluseclayg, and yeah the durable (or anywhere in the mult-phase) was behind a ton of our issues.  The patch I posted just a bit ago that uses the ValueError to map to disconenct causes the proxy to try and things work after a few attempts00:15
claygparse error: causes the *proxy* to *re*-try?!00:15
pelusewell hell someone is... what's the behavior upstream of the obj server on raise ClientDisconnect supposed to be??00:16
*** garthb has quit IRC00:17
claygpeluse: nothing that i'm aware of - i suppose if the proxy looses enough backend connctions it'll return a clean 503 to the client00:17
*** zaitcev has quit IRC00:17
claygpeluse: but those value errors from the object server wouldn't have effected what the proxy does i don't think - the whole point is by the time the object server blows up like that it's because the other end stopped sending bits (or quit)00:18
claygpeluse: it should clean up the logs nice tho!00:18
*** garthb has joined #openstack-swift00:18
*** garthb has quit IRC00:19
*** dmorita has joined #openstack-swift00:19
peluseclayg, well I need to take a much closer look then because what I'm seeing is ssbench reporting 0 errors and a shitload of retries with that patch whereas before it would just show a shitload of errors00:19
*** garthb has joined #openstack-swift00:19
claygFWIW I think swift-bench will retry and log some sort of ... retry ... if it gets a 503 - it might like try every request up to three times - and it counts "unable to make request after three tries" differently that "i retried a few requests here and there"00:19
claygpeluse: interesting!  I think any retries is bad tho :'(00:20
pelusedoesn't it spit back as a 499?00:20
pelusewell, retries when its beacuse of a valid timeout makes a lot of sense I think00:20
claygwell but the client never sees that - or isn't supposed to - all of those errors are "client disconnected before I could send a status"00:20
pelusevalid timeout  meaning a ton of heavy large IO and a 10 sec timeout00:21
pelusesomething is fishy.... but it smells better than before at least :)00:21
*** ajiang has left #openstack-swift00:21
claygpeluse: fair, maybe I misunderstood "a shitload" as more than a percentage point of requests00:22
claygif it's 100K requests and 25 retires it's probably just doing it's job of stressing the system00:22
peluseclayg, yeah its along those lines00:23
claygpeluse: very interesting indeed - I don't too well at accepting stuff on the sleeve tho - i'll try to work through some stuff on that ValueError path tho if it's making an observalbe different for you00:23
peluseclayg, I'll repro and get real numbers though and also look more into why it *appears* that there are no failures with that patch (client view) and there are without it00:23
clayg... i really htought it was just a log cleanup00:23
peluseyeah, I'm not accepting it either.  will see what's happening for real00:24
peluseclayg, well, that's all I thought it was too!00:24
peluse..and then the magic happened :)00:24
claygpeluse: I also want to make sure the tooling isn't giving the wrong impression either - a retry is a filure - the client timed out or got an error and had to retry - the only difference between ss-bench retry and error is just how many times it retried before giving up00:25
clayglol @ magic00:25
*** CR7 has joined #openstack-swift00:28
*** darrenc_afk is now known as darrenc00:29
openstackgerritKota Tsuyuzaki proposed openstack/swift: WIP: PUT/GET path for global ec cluster  https://review.openstack.org/21916500:30
*** rjaiswal has quit IRC00:30
ctennisok clayg peluse the rest of this week for me (aside from my REAL work) is trying to debug a few more reconstruction oddities, hopefully with all outstanding patches merged00:31
ctennisCurrently got one I'm chasing, hoping it turns into nothing.00:31
*** kevinc_ has quit IRC00:34
*** CR7 has quit IRC00:37
*** CR7 has joined #openstack-swift00:38
*** m_kazuhiro has joined #openstack-swift00:38
*** garthb has quit IRC00:39
*** itlinux has joined #openstack-swift00:43
pelusectennis, cool00:45
pelusectennis, we shoudl chat tomorrow about what you're pounding on and what you should expect00:46
ctennissure thing00:46
*** itlinux has quit IRC00:49
*** itlinux has joined #openstack-swift00:50
*** lpabon has joined #openstack-swift01:02
peluseclayg, stil there?01:04
claygpeluse: nope01:04
peluseha!01:04
claygyeah i'm a funny guy01:04
claygpeluse: sup?01:04
*** itlinux has quit IRC01:05
peluseso looking at the buckets patch, not sure whether you or acoles made this change but it appears that we can never hit more than ndata nodes now??01:05
*** kota_ has quit IRC01:05
peluseoh wait, that's the size of the pool.  nevermind :)01:05
claygpeluse: ;)01:05
peluseI haven't had my 3rd beer yet :)01:05
claygpeluse: well you better get on that!01:06
*** lpabon has quit IRC01:07
*** itlinux_ has joined #openstack-swift01:17
*** haomaiwa_ has joined #openstack-swift01:30
*** lcurtis_ has joined #openstack-swift01:36
*** resker has joined #openstack-swift01:48
*** esker has quit IRC01:52
*** nakagawamsa has joined #openstack-swift01:55
*** jkugel has joined #openstack-swift01:59
*** CR7 has quit IRC02:12
*** hrou has joined #openstack-swift02:17
*** itlinux_ has quit IRC02:26
claygpeluse: well you better get on that!02:27
*** ajiang has joined #openstack-swift02:31
*** km has quit IRC02:34
*** resker has quit IRC02:41
*** annegentle has quit IRC02:41
*** annegentle has joined #openstack-swift02:43
*** gyee has quit IRC02:47
*** lpabon has joined #openstack-swift02:53
*** annegentle has quit IRC02:57
*** jkugel has quit IRC03:00
*** changbl has joined #openstack-swift03:02
*** kairo_ has joined #openstack-swift03:05
*** kairo has quit IRC03:07
*** lpabon has quit IRC03:12
*** lpabon has joined #openstack-swift03:15
*** wbhuber has joined #openstack-swift03:15
*** proteusguy_ has quit IRC03:17
*** sanchitmalhotra has joined #openstack-swift03:27
*** haomaiwa_ has quit IRC03:28
*** haomaiwang has joined #openstack-swift03:29
*** proteusguy_ has joined #openstack-swift03:29
*** lpabon has quit IRC03:29
*** esker has joined #openstack-swift03:35
*** esker has quit IRC03:40
*** dimasot has joined #openstack-swift03:42
*** links has joined #openstack-swift03:48
*** wbhuber has quit IRC03:50
*** kei_yama has quit IRC04:06
*** kei_yama has joined #openstack-swift04:06
*** ppai has joined #openstack-swift04:14
*** eranrom has joined #openstack-swift04:16
*** mahatic has joined #openstack-swift04:18
mahaticgood morning04:19
*** lcurtis_ has quit IRC04:22
*** hrou has quit IRC04:22
mattoliveraumahatic: morning!04:23
mahaticmattoliverau: o/04:27
*** esker has joined #openstack-swift04:30
*** esker has quit IRC04:34
*** esker has joined #openstack-swift04:43
*** garthb has joined #openstack-swift04:45
*** trifon has quit IRC04:48
*** esker has quit IRC04:49
*** sanchitmalhotra1 has joined #openstack-swift04:51
*** sanchitmalhotra has quit IRC04:53
*** mtreinish has quit IRC04:57
*** mtreinish has joined #openstack-swift05:04
*** SkyRocknRoll has joined #openstack-swift05:07
*** ppai has quit IRC05:17
*** trifon has joined #openstack-swift05:27
*** ppai has joined #openstack-swift05:31
*** mahatic has quit IRC05:33
*** mahatic has joined #openstack-swift05:34
*** silor has joined #openstack-swift05:36
*** sanchitmalhotra has joined #openstack-swift05:44
*** sanchitmalhotra1 has quit IRC05:46
*** sanchitmalhotra1 has joined #openstack-swift05:49
*** sanchitmalhotra has quit IRC05:52
*** garthb has quit IRC05:52
*** mahatic has quit IRC05:53
*** mahatic_ has joined #openstack-swift05:53
*** sanchitmalhotra has joined #openstack-swift06:06
*** sanchitmalhotra1 has quit IRC06:08
openstackgerritMerged openstack/swift: EC GET path: require fragments to be of same set  https://review.openstack.org/21218706:12
*** chenhuayi has joined #openstack-swift06:37
*** SkyRocknRoll has quit IRC07:04
*** rledisez has joined #openstack-swift07:11
*** nakagawamsa has quit IRC07:13
*** kei_yama has quit IRC07:20
*** kei_yama has joined #openstack-swift07:20
*** vinsh has quit IRC07:34
*** vinsh has joined #openstack-swift07:35
*** ppai has quit IRC07:38
*** esker has joined #openstack-swift07:48
*** geaaru has joined #openstack-swift07:49
*** ppai has joined #openstack-swift07:50
*** eranrom has quit IRC07:52
*** esker has quit IRC07:52
*** eranrom has joined #openstack-swift07:53
*** SkyRocknRoll has joined #openstack-swift08:01
*** hseipp has joined #openstack-swift08:04
*** jistr has joined #openstack-swift08:11
openstackgerritMerged openstack/swift: eventlet.posthook related code needn't anymore  https://review.openstack.org/18389608:13
openstackgerritOpenStack Proposal Bot proposed openstack/swift: Updated from global requirements  https://review.openstack.org/8873608:19
*** [1]dimasot has joined #openstack-swift08:21
*** dimasot has quit IRC08:23
*** [1]dimasot is now known as dimasot08:23
*** SkyRocknRoll has quit IRC08:24
*** ppai has quit IRC08:32
*** marzif has joined #openstack-swift08:43
*** ppai has joined #openstack-swift08:46
*** joeljwright has joined #openstack-swift08:49
*** ChanServ sets mode: +v joeljwright08:49
*** ho_ has joined #openstack-swift08:59
*** ho has quit IRC09:00
*** SkyRocknRoll has joined #openstack-swift09:05
*** kei_yama has quit IRC09:09
*** mahatic_ has quit IRC09:13
*** sanchitmalhotra1 has joined #openstack-swift09:17
*** sanchitmalhotra has quit IRC09:19
openstackgerritBrian Cline proposed openstack/swift: Fixes recon bug with initially missing rings  https://review.openstack.org/21961709:21
*** marzif has quit IRC09:22
openstackgerritBrian Cline proposed openstack/swift: Fixes recon bug with initially missing rings  https://review.openstack.org/21961709:24
*** eranrom has quit IRC09:25
*** m_kazuhiro has quit IRC09:36
*** esker has joined #openstack-swift09:36
*** silor has quit IRC09:37
*** esker has quit IRC09:41
*** wbhuber has joined #openstack-swift09:46
*** wbhuber has quit IRC09:51
*** ppai has quit IRC10:00
*** marzif has joined #openstack-swift10:06
*** haypo has joined #openstack-swift10:07
haypohi. congrats for the release 2.4 ;)10:09
haypoi'm back for python3! i splitted my patches for fix "tox -e py34" into smaller patches, easier to review:10:09
haypohttps://review.openstack.org/#/c/217423/ py3: Update pbr and dnspython requirements10:10
haypohttps://review.openstack.org/#/c/217424/ Input validation must not depend on the locale10:10
haypohttps://review.openstack.org/#/c/217425/ Port swob to Python 310:10
haypohttps://review.openstack.org/#/c/217426/ py3: Fix Python 3 issues in utils10:10
haypohttps://review.openstack.org/#/c/199034/ py3: Add py34 test environment to tox10:10
*** esker has joined #openstack-swift10:14
*** ppai has joined #openstack-swift10:14
*** silor has joined #openstack-swift10:16
*** acoles_ is now known as acoles10:18
*** esker has quit IRC10:18
acolespeluse: clayg: yay etag buckets merged :)10:25
*** slo_ has joined #openstack-swift10:28
*** slo_ has quit IRC10:29
*** slo_ has joined #openstack-swift10:29
*** slo_ has quit IRC10:29
*** chmouel has quit IRC10:40
*** chmouel has joined #openstack-swift10:41
*** sanchitmalhotra has joined #openstack-swift10:46
*** sanchitmalhotra1 has quit IRC10:48
*** daemontool_ has joined #openstack-swift10:50
*** esker has joined #openstack-swift10:50
openstackgerritChristian Schwede proposed openstack/swift: Test if container_sweep is executed on unmounted devices  https://review.openstack.org/21963910:55
*** wbhuber has joined #openstack-swift11:01
*** chenhuayi has quit IRC11:02
*** mlanner has quit IRC11:02
*** chrisnelson has quit IRC11:02
*** chrisnelson has joined #openstack-swift11:03
*** joeljwright has quit IRC11:04
*** mlanner has joined #openstack-swift11:04
*** dmorita has quit IRC11:05
*** hseipp has quit IRC11:07
openstackgerritMerged openstack/swift: Fix dispersion-reports error message  https://review.openstack.org/21369711:12
*** Anticimex has quit IRC11:13
*** sanchitmalhotra1 has joined #openstack-swift11:17
*** sanchitmalhotra has quit IRC11:19
*** pbourke has joined #openstack-swift11:22
pbourkehi - could anyone tell me in short how are the swift replicator processes and rsync related?11:23
pbourkeor are they?11:23
*** Anticimex has joined #openstack-swift11:23
*** esker has quit IRC11:26
ctennispbourke: the replicator process (when using the rsync replication method) handles contacting the remote nodes to determine if some form of replication is needed, and if so, starts rsync to actually do the replication11:26
*** m_kazuhiro has joined #openstack-swift11:26
*** m_kazuhiro has quit IRC11:26
ctennisin other words, rsync does the actual data transfer, but the logic is handled by the replicator11:26
pbourkectennis: that's kind of my understanding, thanks11:26
*** esker has joined #openstack-swift11:26
pbourkectennis: is there a reason docs seem to start rsync as a daemon then?11:26
ctenniswell, rsync listens as a daemon on the remote side so that the rsync client can send data to it11:27
pbourkegotcha11:27
ctennisit doesn't connect over ssh like you might be used to11:27
*** hrou has joined #openstack-swift11:32
*** pbourke has quit IRC11:36
*** pbourke has joined #openstack-swift11:36
*** aix has quit IRC11:36
*** haomaiwang has quit IRC11:37
*** joeljwright has joined #openstack-swift11:38
*** ChanServ sets mode: +v joeljwright11:38
*** ctrath has joined #openstack-swift11:42
*** bapalm has quit IRC11:43
*** bapalm has joined #openstack-swift11:49
*** breitz1 has quit IRC11:51
*** breitz has joined #openstack-swift11:51
*** esker has quit IRC11:52
*** dimasot has quit IRC11:54
*** ppai has quit IRC12:01
*** lpabon has joined #openstack-swift12:03
*** petertr7_away is now known as petertr712:03
*** aix has joined #openstack-swift12:09
openstackgerritAlistair Coles proposed openstack/python-swiftclient: Cleanup and improve tests for download  https://review.openstack.org/21724612:10
*** ppai has joined #openstack-swift12:14
*** delattec has quit IRC12:15
*** cdelatte has quit IRC12:15
*** petertr7 is now known as petertr7_away12:17
*** peluse has quit IRC12:22
*** peluse has joined #openstack-swift12:23
*** petertr7_away is now known as petertr712:23
*** ChanServ sets mode: +v peluse12:24
*** bkopilov has quit IRC12:25
*** bkopilov has joined #openstack-swift12:26
openstackgerritMerged openstack/swift: Included reference in swift.obj.diskfile to enumerate the string used for data file paths.  https://review.openstack.org/21943812:27
*** lpabon has quit IRC12:27
*** david-lyle has quit IRC12:27
*** mordred has quit IRC12:27
openstackgerritMerged openstack/swift: Fix EC range GET/COPY handling  https://review.openstack.org/20489212:27
openstackgerritMerged openstack/swift: Preserve traceback in swift-dispersion-report  https://review.openstack.org/21947112:27
*** lpabon has joined #openstack-swift12:27
*** david-lyle has joined #openstack-swift12:28
*** ctrath has quit IRC12:29
*** peluse has quit IRC12:29
*** peluse has joined #openstack-swift12:35
*** ChanServ sets mode: +v peluse12:35
*** peluse_ has joined #openstack-swift12:37
*** peluse has quit IRC12:38
*** ppai has quit IRC12:39
*** peluse has joined #openstack-swift12:43
*** ChanServ sets mode: +v peluse12:43
peluse_test test test12:45
pelusethis is also a test12:45
acolespeluse: this isn't :)12:46
*** janonymous_ has joined #openstack-swift12:46
pelusetrying new IRC client with my new Mac... man these things can be vastly different from one to another...12:47
ctennispeluse: TOO EARLY12:47
*** NM has joined #openstack-swift12:48
*** annegentle has joined #openstack-swift12:51
*** itlinux_ has joined #openstack-swift12:51
*** peluse_ has quit IRC12:52
*** peluse_ has joined #openstack-swift12:52
pelusectennis: its never too early :)12:53
pelusectennis: which do you use?12:53
*** peluse_ has quit IRC12:53
ctennisI use Linkinus 2, but it hasn't seen updates in a long time and is pretty shoddy.12:54
pelusectennis: was just reading about that one, trying limechat now12:54
ctennisI wouldn't recommend it starting new, I'm just used to it by now.12:54
tdasilvanotmyname: just got the invite for Openstack Meetup in Boston this month and look who is speaking: http://www.meetup.com/Openstack-Boston/events/22505740012:55
janonymous_Share slides after this please :)12:57
*** lpabon has quit IRC13:01
*** szaher has joined #openstack-swift13:06
*** bill_az has joined #openstack-swift13:12
*** links has quit IRC13:17
*** esker has joined #openstack-swift13:18
*** esker has quit IRC13:20
*** nadeem has joined #openstack-swift13:25
*** ho_ has quit IRC13:25
*** esker has joined #openstack-swift13:25
*** esker has quit IRC13:26
*** jkugel has joined #openstack-swift13:26
*** esker has joined #openstack-swift13:27
*** zul has quit IRC13:29
*** zul has joined #openstack-swift13:32
*** dustins has joined #openstack-swift13:36
*** thurloat_isgone is now known as thurloat13:38
*** itlinux_ has quit IRC13:43
*** annegentle has quit IRC13:43
openstackgerritjanonymous proposed openstack/swift: Use six.moves.map/filter/zip inplace of itertools.izip/imap/ifilter  https://review.openstack.org/20341413:57
*** haomaiwang has joined #openstack-swift13:57
*** petertr7 is now known as petertr7_away14:02
*** trifon has quit IRC14:02
*** sanchitmalhotra1 has quit IRC14:03
*** peluse has quit IRC14:03
*** wbhuber has quit IRC14:04
*** petertr7_away is now known as petertr714:05
*** marzif has quit IRC14:05
*** marzif has joined #openstack-swift14:06
*** peluse has joined #openstack-swift14:07
*** xnox has quit IRC14:07
*** xnox has joined #openstack-swift14:09
*** peluse has quit IRC14:11
*** janonymous_ has quit IRC14:14
*** annegentle has joined #openstack-swift14:14
*** jrichli has joined #openstack-swift14:16
*** haomaiwang has quit IRC14:23
*** haomaiwa_ has joined #openstack-swift14:25
*** cdelatte has joined #openstack-swift14:25
*** delattec has joined #openstack-swift14:25
*** dustins_ has joined #openstack-swift14:26
*** esker has quit IRC14:27
*** dustins has quit IRC14:28
*** mahatic has joined #openstack-swift14:31
*** kairo_ has quit IRC14:37
*** kairo has joined #openstack-swift14:38
*** esker has joined #openstack-swift14:39
*** wbhuber has joined #openstack-swift14:39
*** petertr7 is now known as petertr7_away14:42
*** wshao has joined #openstack-swift14:44
*** cdelatte has quit IRC14:47
*** delattec has quit IRC14:47
*** cdelatte has joined #openstack-swift14:47
*** delattec has joined #openstack-swift14:47
*** jlhinson has joined #openstack-swift14:50
*** esker has quit IRC14:53
*** peluse has joined #openstack-swift14:54
*** ctrath has joined #openstack-swift14:58
*** ctrath has quit IRC14:58
*** minwoob has joined #openstack-swift15:07
*** jistr is now known as jistr|call15:12
aerwin3Good Morning everyone.15:16
pelusegood morning15:18
*** janonymous_ has joined #openstack-swift15:22
*** ctrath has joined #openstack-swift15:23
*** chsc has joined #openstack-swift15:24
*** dustins_ has quit IRC15:25
openstackgerritjanonymous proposed openstack/swift: Python 3 using builtins instead of __builtin__, rename raw_input() to input() from six.moves  https://review.openstack.org/20093215:27
*** wshao has quit IRC15:30
*** wshao has joined #openstack-swift15:30
*** szaher has quit IRC15:34
*** peluse_ has joined #openstack-swift15:36
*** peluse_ has quit IRC15:36
*** jistr|call is now known as jistr15:38
*** wshao has quit IRC15:45
*** hseipp has joined #openstack-swift15:47
*** hseipp has quit IRC15:47
*** hseipp has joined #openstack-swift15:47
notmynamegood morning15:49
*** trifon has joined #openstack-swift15:49
*** dustins has joined #openstack-swift15:49
*** dustins has quit IRC15:55
notmynameFYI (and i'll be raising this again in today's meeting), I'l looking at the end of this month (ie 4 weeks from today) for the next swift release. the one in the overall Liberty release15:56
notmynameI'd like to have all the known EC bugs closed for that. That is my focus for the next release15:56
*** esker has joined #openstack-swift15:57
*** dustins has joined #openstack-swift15:57
*** gyee has joined #openstack-swift15:57
notmynamealso, kota_ has raised his global EC spec for the meeting. please look at the links on https://wiki.openstack.org/wiki/Meetings/Swift to prepare for the discussion15:59
*** esker has quit IRC15:59
hayponotmyname: hi. would it be possible to put python3 somewhere in your agenda? :)16:00
haypoby the way, does someone know if PyEClib 1.0.8 is going to be released or not?16:02
haypoi was tagged almost one month ago16:02
notmynamehaypo: no for py316:03
notmynameit's not something that can be completed by the next release, so I want to prioritize those things that can be completed16:03
hayponotmyname: ok16:04
*** petertr7_away is now known as petertr716:04
*** nadeem has quit IRC16:04
*** garthb has joined #openstack-swift16:07
*** esker has joined #openstack-swift16:14
*** devlaps has joined #openstack-swift16:16
openstackgerritAlistair Coles proposed openstack/swift: Merge branch 'master' into feature/crypto  https://review.openstack.org/21977516:17
wbhubernotmyname: for the next swift release (2.4), is 9/30 the last day to deliver and review EC fixes before merging in time for Liberty?16:17
acolesjrichli: ^^16:17
notmynamewbhuber: next swift release will be either 2.4.1 or 2.5.0 (depending on what lands)16:17
notmynamewbhuber: but the patches need to be landed by the end of the month (to within a few days)16:18
wbhubernotmyname: good to know16:18
*** petertr7 is now known as petertr7_away16:21
*** esker has quit IRC16:23
*** rledisez has quit IRC16:25
*** doffm has left #openstack-swift16:26
*** itlinux has joined #openstack-swift16:31
itlinuxquick question do we have have a wild card for swift download container *.txt example..16:31
itlinuxthanks16:31
*** nadeem has joined #openstack-swift16:33
*** haypo has left #openstack-swift16:34
peluseBTW update on the EC perf cluster at Intel; all tests ran w/no EC realted issues that I tell.  One test failed with a bunch of 401s on container create - started it over again last night and will check on it again in a few hrs.  Running master + one pending patch 21133816:42
patchbotpeluse: https://review.openstack.org/#/c/211338/16:42
*** marzif has quit IRC16:43
*** marzif has joined #openstack-swift16:43
*** jistr has quit IRC16:44
*** nadeem has quit IRC16:44
*** zhill_desktop has joined #openstack-swift16:47
*** hrou has quit IRC16:49
ctennispeluse: let me know when you want to chat16:57
*** SkyRocknRoll has quit IRC16:57
*** peluse has quit IRC16:58
*** hseipp has quit IRC16:59
*** annegentle has quit IRC17:01
*** lcurtis has joined #openstack-swift17:06
*** peluse has joined #openstack-swift17:08
*** peluse has quit IRC17:12
*** wshao_ has joined #openstack-swift17:17
*** janonymous_ has quit IRC17:18
*** marzif has quit IRC17:20
*** jlhinson has quit IRC17:24
*** mahatic has quit IRC17:25
*** mahatic has joined #openstack-swift17:25
*** wshao_ has quit IRC17:26
*** dimasot has joined #openstack-swift17:35
*** acoles is now known as acoles_17:37
*** [1]dimasot has joined #openstack-swift17:38
*** wshao_ has joined #openstack-swift17:39
*** aix has quit IRC17:40
*** dimasot has quit IRC17:40
*** [1]dimasot is now known as dimasot17:40
*** thurloat is now known as thurloat_isgone17:44
openstackgerritSamuel Merritt proposed openstack/swift: Add functional test for multi-range GET requests.  https://review.openstack.org/16782817:48
*** nadeem has joined #openstack-swift17:54
*** nadeem has quit IRC17:54
*** thurloat_isgone is now known as thurloat17:54
*** nadeem has joined #openstack-swift17:55
*** dustins has quit IRC17:56
*** [1]dimasot has joined #openstack-swift17:59
*** dustins has joined #openstack-swift17:59
jrichliacoles_: thanks!18:00
*** dimasot has quit IRC18:02
*** [1]dimasot is now known as dimasot18:02
claygacoles_: thanks for the review on patch 21802318:03
patchbotclayg: https://review.openstack.org/#/c/218023/18:03
*** geaaru has quit IRC18:04
claygcommunity service annoucements: out-of-date doc strings are BUGS - in some cases one of the worst kinda because it'll only hurt your friendly neighborhood swift contributor - and if there's any one we like more than operators - it's ourselves - generally we shouldn't merge bugs if can avoid it18:05
notmynameI suspect that comment has a funny story behind it...18:06
claygnot really - just acoles_ being too nice with me being sloppy with a docstring18:07
*** bill_az_ has joined #openstack-swift18:07
claygI mean he could have just pushed over - but w/e18:07
claygi have it checked out now18:07
*** marzif has joined #openstack-swift18:09
claygp.s. I hate paste.openstack.org - #1 it uses drizzle #2 it's slow #3 it's raw output strips the trailing newline that you need to be able to curl ... | git apply18:11
claygI try  --whitespace fix  I try  --inaccurate-eof - but always it hates me18:11
claygeventually I result to git apply copy paste body of patch hit enter ctrl-D18:11
claygthen I rant into irc for like 5 lines18:12
*** jlhinson has joined #openstack-swift18:19
*** itlinux has quit IRC18:19
*** annegentle has joined #openstack-swift18:20
*** marzif has quit IRC18:22
*** marzif has joined #openstack-swift18:22
*** mahatic has quit IRC18:23
*** mahatic has joined #openstack-swift18:24
*** marzif has quit IRC18:28
*** aix has joined #openstack-swift18:34
*** esker has joined #openstack-swift18:34
brianclineheh wow, my local env didn't barf the way jenkins did on my recon patch...my local env seems busted18:34
*** esker has quit IRC18:35
*** esker has joined #openstack-swift18:35
*** petertr7_away is now known as petertr718:38
notmynameall, charz has just finished setting up probe tests on the community CI cluster. so now every patch will have the probe tests run against them18:38
notmynamethe first result can be seen at https://review.openstack.org/#/c/167828/18:38
notmynamethanks charz!!18:38
charznotmyname: np :)18:38
*** zaitcev has joined #openstack-swift18:38
*** ChanServ sets mode: +v zaitcev18:38
*** silor has quit IRC18:39
*** peluse has joined #openstack-swift18:56
*** mahatic has quit IRC19:08
*** hrou has joined #openstack-swift19:13
*** annegentle has quit IRC19:17
*** bill_az_ has quit IRC19:23
ctennispeluse: clayg: still running into https://bugs.launchpad.net/swift/+bug/1489546 even with patch applied19:27
openstackLaunchpad bug 1489546 in OpenStack Object Storage (swift) "logic error in ssync_rcvr when getting EC frags from a handoff" [High,In progress] - Assigned to paul luse (paul-e-luse)19:27
ctenniserr standby, may be user error19:30
*** wshao_ has quit IRC19:30
*** garthb_ has joined #openstack-swift19:33
*** garthb has quit IRC19:35
*** annegentle has joined #openstack-swift19:35
*** gyee has quit IRC19:44
ctennisdefinitely user error19:48
*** esker has quit IRC19:48
openstackgerritJonathan Hinson proposed openstack/swift: Automatic refresh of memcache config settings  https://review.openstack.org/21849019:52
*** ctrath has quit IRC19:55
*** ctrath has joined #openstack-swift19:58
*** garthb_ has quit IRC20:02
*** garthb_ has joined #openstack-swift20:02
*** bill_az_ has joined #openstack-swift20:03
*** dimasot has quit IRC20:05
*** gyee has joined #openstack-swift20:06
albertomswift 2.4.0 ?20:10
albertomis this supposed to be liberty or kilo ?20:10
ctennisneither, swift usually has a midcycle release as well.20:11
albertomo.O20:12
*** petertr7 is now known as petertr7_away20:12
albertomso is it safe to use swift 2.4.0 with the remaining components in kilo (2015.1.1)20:15
albertomsafe/recommended20:15
ctennisthere's no reason it wouldn't work, it's always been backwards compatible20:20
*** eranrom has joined #openstack-swift20:20
* albertom will give it a try20:21
*** nadeem has quit IRC20:23
eranromGreetings. Came to poke a clayg regarding a certain review :-)20:24
ctennisthere's also no harm just running the kilo version if you don't need the explicit new features20:24
albertomctennis: where can i find a release notes for this new 2.4.0 ?20:26
ctennishttp://permalink.gmane.org/gmane.comp.cloud.openstack.devel/6295920:26
*** hrou has quit IRC20:30
*** zhill has joined #openstack-swift20:44
*** wshao has joined #openstack-swift20:46
notmynamemeeting in 11 minutes20:49
notmyname(all meetings, all day for me)20:49
*** ho has joined #openstack-swift20:50
notmynamealbertom: yeah, 2.4.0 should work fine with any other openstack components. also, we'll probably have another release in about a month20:50
jrichlinotmyname: but this one will be the most enjoyable, right ;-)20:50
notmynamejrichli: let's hope :-)20:50
*** wshao has quit IRC20:51
hogood morning!20:53
notmynamehello ho20:53
honotmyname: hi!20:54
*** kota_ has joined #openstack-swift20:56
*** ChanServ sets mode: +v kota_20:56
kota_good morning20:56
hokota_: morning!20:56
kota_ho: morning^^20:57
mattoliveraumorning20:58
homattoliverau: morning!20:58
*** acoles_ is now known as acoles20:58
notmynameswift meeting time in #openstack-meeting20:58
*** ctrath has quit IRC20:59
notmynamemattoliverau: looks like I'll be going to geelong21:00
mattoliveraunotmyname: awesome!! \o/21:01
*** ctrath has joined #openstack-swift21:02
*** petertr7_away is now known as petertr721:04
acolesclayg: whats this about me being too nice?21:04
claygwee!21:05
*** eranrom has quit IRC21:07
*** ctrath has quit IRC21:08
*** ctrath has joined #openstack-swift21:09
*** ctrath has quit IRC21:11
*** petertr7 is now known as petertr7_away21:19
*** trifon has quit IRC21:19
*** petertr7_away is now known as petertr721:20
*** ctrath has joined #openstack-swift21:25
tdasilvacharz: thanks for the probe tests work!21:30
acolescharz: thanks for enabling probe tests on the qa cluster21:30
mattoliveraunice work charz!21:30
*** marzif has joined #openstack-swift21:44
flwang1cschwede: ping21:44
*** legionarius has joined #openstack-swift21:45
*** petertr7 is now known as petertr7_away21:46
*** bill_az_ has quit IRC21:49
openstackgerritMerged openstack/swift-specs: Add global ec cluster improvement  https://review.openstack.org/20944721:52
claygwell that was nice21:53
*** legionarius has left #openstack-swift21:53
kota_nice! thanks everyone for landing that!21:54
notmynameahale: thanks for the review21:54
*** joeljwright1 has joined #openstack-swift21:54
notmynameahale: (on the graphite patch)21:54
claygahale: does reviews!?  that's awesome!21:55
notmynamehe just did!21:55
claygdfg_: redbo: hurricanerix: ^ see it's not so hard21:55
*** bill_az_ has joined #openstack-swift21:55
ahaleno worries, I was just remembering the pain we had when we added proxy name in some of our metrics and it all shifted up one :)21:55
mattoliveraulol21:55
*** joeljwright has quit IRC21:57
hurricanerixclayg: every time i look for a review, it already has so many comments on it, there isn't anything new for me to say.  But I will keep looking =)21:58
*** bill_az has quit IRC21:58
* mattoliverau goes to find breakfast21:59
clayghurricanerix: that can't be correct - there's a whole field in the dashboard for "hasn't had a review comment"21:59
torgomaticahale: at least this one's purely additive, so you don't really have to rename anything22:00
torgomatic(which, you're right, *is* a giant pain in the rear)22:00
bill_az_also, I think there may be others who attend - peluse mentioned Intel wanted to start participating22:01
*** bkopilov has quit IRC22:01
bill_az_sorry - last one was meant for swift3 channel ;-(22:01
kota_bill_az_: no worries22:01
hurricanerixclayg: I should say every time I see something I think is interesting to review, I end up having nothing to add to it. =)22:02
*** wshao has joined #openstack-swift22:02
ahaleyeah torgomatic , its not the end of the world at all, just not something you'd want as a surprise :)22:02
*** wshao has quit IRC22:03
torgomatictrue :)22:03
*** dustins has quit IRC22:03
*** acoles is now known as acoles_22:05
*** bkopilov has joined #openstack-swift22:06
openstackgerritClay Gerrard proposed openstack/swift: Fix purge for tombstone only REVERT job  https://review.openstack.org/21802322:07
*** annegentle has quit IRC22:08
*** NM has quit IRC22:09
*** jkugel has quit IRC22:10
claygnotmyname: torgomatic: acoles_: so i'm looking at get_info for unrelated reasons and remmebering that in the process of container-tempurl fixes we had like 3 possible cleanups to get_info which were all very good - we should idk what we should do acctually - but it was useful code!22:10
*** chsc has quit IRC22:13
*** thurloat is now known as thurloat_isgone22:13
*** marzif has quit IRC22:18
redboclayg: it's hard to find things to care about!  Once you skip EC and python 3, there's not much there.22:20
claygconcurrent gets22:20
claygi didn't even have to think about that one - i'm sure there's more - do you really want me to help find you some?  (I don't want to waste my time)22:21
*** zhill has quit IRC22:21
redboI reviewed the posthook thing.  It was removing code, i loved it.22:22
claygnice!22:22
notmynameheh22:22
*** bkopilov has quit IRC22:22
claygnotmyname: those little gerrit bar color twiddles would have been sweet - I don't know why people were fussig about them22:23
notmynameclayg: I've seen that some people still have them. mostly -infra people22:23
notmynamedont' know how22:23
*** AndreiaKumpera has quit IRC22:24
notmynameencryption. container sharding. tons in the client. cluster full docs/features. slorange. metadata indexing. increase part power. fast post. container sync. reverse listings. better ring deployment. per-drive rsync. policy descriptions. utility to undelete accounts. copy with ranges. more atomic object puts. more expressive ACLs22:27
*** bkopilov has joined #openstack-swift22:28
*** garthb__ has joined #openstack-swift22:30
*** joeljwright1 has quit IRC22:31
*** garthb_ has quit IRC22:32
openstackgerritRomain LE DISEZ proposed openstack/swift: Allows to configure the rsync modules where the replicators will send data  https://review.openstack.org/19218622:36
*** lcurtis has quit IRC22:36
openstackgerritHisashi Osanai proposed openstack/swift: Add functional test for access control (PUT) with Keystone  https://review.openstack.org/20241122:37
*** zhill has joined #openstack-swift22:38
*** thurloat_isgone is now known as thurloat22:42
*** haomaiwang has joined #openstack-swift22:43
*** haomaiwa_ has quit IRC22:46
claygOMG this patch is *so* great -> https://review.openstack.org/#/c/192186/322:48
claygi feel like I joined the make a wish foundation22:48
*** thurloat is now known as thurloat_isgone22:51
*** thurloat_isgone is now known as thurloat22:51
notmynamersync per device. nice22:52
*** jlhinson has quit IRC22:59
*** ctrath has quit IRC23:01
*** chsc has joined #openstack-swift23:03
*** km has joined #openstack-swift23:05
*** jrichli has quit IRC23:06
*** zaitcev has quit IRC23:12
openstackgerritMerged openstack/swift: Adds useful dispersion info from changelog  https://review.openstack.org/21932223:14
*** hrou has joined #openstack-swift23:17
*** minwoob has quit IRC23:21
*** AndreiaKumpera has joined #openstack-swift23:21
*** thurloat is now known as thurloat_isgone23:24
*** kei_yama has joined #openstack-swift23:30
openstackgerritSamuel Merritt proposed openstack/swift: Add functional test for multi-range GET requests.  https://review.openstack.org/16782823:31
*** wbhuber has quit IRC23:33
*** wshao has joined #openstack-swift23:34
*** wshao has quit IRC23:35
*** thurloat_isgone is now known as thurloat23:37
claygOMG this is so awesome -> https://8b86aea46fb38e6450f2-0e5f4c086da474abc1df58826577db2f.ssl.cf1.rackcdn.com/192186/16/probetests/result.html23:39
claygprobetests for the rsync per device setup!?  it's like... why do we even need reviews anymore :P23:39
openstackgerritOpenStack Proposal Bot proposed openstack/swift: Updated from global requirements  https://review.openstack.org/8873623:50
*** thurloat is now known as thurloat_isgone23:50
*** wshao has joined #openstack-swift23:53
timburkeclayg: just start +2ing everything! what could go wrong?23:53
mattoliveraunice :)23:54
*** wshao has quit IRC23:57

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!