Thursday, 2017-08-17

*** lucasxu has joined #openstack-swift00:18
*** lucasxu has quit IRC00:19
*** wasmum has quit IRC00:21
*** kiennt has joined #openstack-swift00:47
mattoliveraukota_: congrats on becoming the storlets ptl!00:50
*** cshastri has joined #openstack-swift00:52
*** baojg has quit IRC00:56
*** shuyingya has joined #openstack-swift01:05
*** shuyingya has quit IRC01:09
*** klrmn has quit IRC01:12
*** Sukhdev has quit IRC01:18
*** shuyingya has joined #openstack-swift01:28
*** vint_bra has joined #openstack-swift01:35
*** fanzhang has quit IRC01:41
*** fanzhang has joined #openstack-swift01:42
*** itlinux has joined #openstack-swift01:43
*** vint_bra has quit IRC01:46
*** mat128 has joined #openstack-swift01:53
*** baojg has joined #openstack-swift01:57
*** baojg has quit IRC02:02
*** itlinux has quit IRC02:15
*** itlinux has joined #openstack-swift02:15
*** itlinux has quit IRC02:17
*** itlinux has joined #openstack-swift02:18
*** vint_bra has joined #openstack-swift02:26
*** klrmn has joined #openstack-swift02:39
*** vint_bra has quit IRC02:41
openstackgerritMerged openstack/swift master: Always check for unexpected requests in mocked_http_conn  https://review.openstack.org/46148302:52
openstackgerritMerged openstack/swift master: Add test for all-nodes-error-limited  https://review.openstack.org/49309402:54
*** mat128 has quit IRC03:03
*** Sukhdev has joined #openstack-swift03:19
*** kei_yama has quit IRC03:34
*** kei_yama has joined #openstack-swift03:43
*** gyee has quit IRC03:44
*** Sukhdev has quit IRC03:57
*** gkadam has joined #openstack-swift03:58
*** baojg has joined #openstack-swift03:59
*** gkadam_ has joined #openstack-swift04:01
*** links has joined #openstack-swift04:01
*** gkadam has quit IRC04:03
*** baojg has quit IRC04:03
*** gkadam_ is now known as gkadam04:04
*** silor has joined #openstack-swift04:06
*** silor has quit IRC04:14
*** openstack has quit IRC04:42
*** openstack has joined #openstack-swift04:45
*** pdardeau has joined #openstack-swift04:45
*** cebruns_ has joined #openstack-swift04:45
*** csmart has quit IRC04:45
*** stradling has joined #openstack-swift04:45
*** mtreinish has joined #openstack-swift04:45
*** sai has joined #openstack-swift04:45
*** Dinesh_Bhor has joined #openstack-swift04:45
*** fanzhang has joined #openstack-swift04:45
*** larivee has joined #openstack-swift04:46
*** zigo has joined #openstack-swift04:46
*** nikivi has joined #openstack-swift04:46
*** openstackstatus has joined #openstack-swift04:46
*** mattoliverau has joined #openstack-swift04:46
*** timburke has joined #openstack-swift04:46
*** ejat has joined #openstack-swift04:46
*** EmilienM has joined #openstack-swift04:46
*** shuyingya has joined #openstack-swift04:46
*** ChanServ sets mode: +v openstackstatus04:46
*** okdas has joined #openstack-swift04:46
*** kei_yama has joined #openstack-swift04:46
*** dancn has joined #openstack-swift04:46
*** ejat has quit IRC04:46
*** ejat has joined #openstack-swift04:46
*** ChanServ sets mode: +v timburke04:47
*** ujjain has quit IRC04:47
*** ujjain has joined #openstack-swift04:47
*** etienneme has joined #openstack-swift04:47
*** hugokuo has joined #openstack-swift04:47
*** okdas has quit IRC04:47
*** okdas has joined #openstack-swift04:47
*** neonpastor has joined #openstack-swift04:47
*** kota_ has joined #openstack-swift04:48
*** htruta has joined #openstack-swift04:48
*** ChanServ sets mode: +v kota_04:48
*** Sukhdev has joined #openstack-swift04:48
*** charz has joined #openstack-swift04:48
*** jistr has joined #openstack-swift04:49
*** glb has joined #openstack-swift04:49
*** jaosorior has joined #openstack-swift04:49
*** Jeffrey4l has joined #openstack-swift04:49
*** notmyname has joined #openstack-swift04:50
*** ahale has joined #openstack-swift04:50
*** ChanServ sets mode: +v notmyname04:50
*** csmart has joined #openstack-swift04:50
*** jcook has joined #openstack-swift04:50
*** jidar has joined #openstack-swift04:51
*** cshastri has joined #openstack-swift04:52
*** raginbajin has joined #openstack-swift04:55
*** tristanC has joined #openstack-swift04:56
*** kiennt has joined #openstack-swift04:59
*** szaher has joined #openstack-swift05:04
*** itlinux has quit IRC05:11
*** itlinux has joined #openstack-swift05:31
*** jaosorior has quit IRC05:32
*** jaosorior has joined #openstack-swift05:33
*** itlinux has quit IRC05:41
*** kiennt has quit IRC05:41
*** mvk has joined #openstack-swift05:48
*** stewie925 has joined #openstack-swift05:49
*** Sukhdev has quit IRC05:53
kota_mattoliverau: oh, you know! I don't think the tc decision would be done yet, right? :P05:55
mattoliverauthe email came out this morning :)05:56
mattoliveraukota_: ^05:56
kota_mattoliverau: oh really, i missed it. For openstack-dev mailing list?05:56
*** mvk_ has joined #openstack-swift05:57
*** rcernin has joined #openstack-swift05:57
mattoliveraukota_: http://lists.openstack.org/pipermail/openstack-dev/2017-August/121206.html05:58
kota_mattoliverau: thanks, i just found it :)05:58
*** mvk has quit IRC05:59
kota_but patch 493846 looks still under waiting decision :/05:59
kota_patch bot?06:00
*** maestropandy has joined #openstack-swift06:00
kota_https://review.openstack.org/#/c/49384606:00
*** maestropandy has left #openstack-swift06:00
kota_hmm... patch bot is absent today06:01
kota_perhaps he is in vacation :P06:01
*** kiennt has joined #openstack-swift06:09
*** openstackgerrit has joined #openstack-swift06:22
openstackgerritMerged openstack/swift master: Fix deadlock when logging from a tpool thread.  https://review.openstack.org/49363606:22
*** kiennt has quit IRC06:22
*** tovin07_ has joined #openstack-swift06:22
*** ianychoi has quit IRC06:23
*** kiennt has joined #openstack-swift06:23
kota_oh, that deadlock fix got merged, nice.06:23
*** saltsa has joined #openstack-swift06:25
*** ianychoi has joined #openstack-swift06:26
*** geaaru has joined #openstack-swift06:33
*** hseipp has joined #openstack-swift06:41
*** pcaruana has joined #openstack-swift06:43
*** tesseract has joined #openstack-swift06:47
*** hoonetorg has joined #openstack-swift06:54
*** Jeffrey4l has quit IRC06:56
*** Jeffrey4l has joined #openstack-swift06:57
*** baojg has joined #openstack-swift07:01
*** baojg has quit IRC07:05
*** skudlik has joined #openstack-swift07:08
*** skudlik has left #openstack-swift07:08
*** Dinesh_Bhor has quit IRC07:11
*** Dinesh_Bhor has joined #openstack-swift07:18
*** bob_cheesey1 has joined #openstack-swift07:23
*** bob_cheesey1 has joined #openstack-swift07:24
*** bob_cheesey1 has quit IRC07:25
*** gkadam has joined #openstack-swift07:26
acolesgood morning07:28
*** mkrcmari__ has joined #openstack-swift07:28
*** mvk_ has quit IRC07:32
fanzhangacoles: good afternoon :)07:32
*** mkrcmari__ has quit IRC07:33
*** mathiasb has joined #openstack-swift07:34
*** cbartz has joined #openstack-swift07:36
*** tovin07 has joined #openstack-swift07:45
*** mkrcmari__ has joined #openstack-swift07:46
*** tesseract has quit IRC07:58
*** oshritf has joined #openstack-swift07:59
*** hseipp has quit IRC08:00
*** openstackgerrit has quit IRC08:17
*** openstackgerrit has joined #openstack-swift08:18
openstackgerritjunbo.li proposed openstack/swift master: Unsupported 'message' Exception attribute in PY3  https://review.openstack.org/48701108:18
*** Anticimex has joined #openstack-swift08:19
*** joeljwright has joined #openstack-swift08:26
*** ChanServ sets mode: +v joeljwright08:26
kota_good morning acoles08:42
*** fanzhang_ has joined #openstack-swift08:53
*** fanzhang has quit IRC08:56
*** remix_tj has joined #openstack-swift09:10
acoleskota_: o/09:10
joeljwrightacoles: kota_: good morning09:10
*** fanzhang has joined #openstack-swift09:11
kota_good morning joeljwright09:11
acolesjoeljwright: good morning09:11
*** fanzhang_ has quit IRC09:12
*** afazekas is now known as afazekas|sick09:19
openstackgerritMerged openstack/swift master: Remove unnecessary exception handling  https://review.openstack.org/48442809:31
*** gkadam is now known as gkadam-afk09:32
*** afazekas|sick is now known as afazekas09:47
*** kei_yama has quit IRC09:55
*** mvk_ has joined #openstack-swift10:04
*** mkrcmari__ has quit IRC10:08
*** kong has quit IRC10:10
*** janonymous has quit IRC10:11
*** kong has joined #openstack-swift10:13
*** skudlik has joined #openstack-swift10:13
*** janonymous has joined #openstack-swift10:15
kota_cbartz: here?10:21
*** kiennt has quit IRC10:24
openstackgerritjunbo.li proposed openstack/swift master: Unsupported 'message' Exception attribute in PY3  https://review.openstack.org/48701110:44
cbartzkota_: Yes10:52
cbartzI am just reading your newest comment10:53
kota_hello cbartz10:54
kota_yeah, just pushed it out10:55
kota_cbartz: I mainly concern user affection, right now on the patch. I think quota is not only reference when the used-bytes exceeded but the capacity planning for users10:57
kota_the idea for the charge factor sounds interesting but it could be complicated for users and it seems not directly related to the quota mechanism10:58
*** cshastri has quit IRC10:59
kota_and the use case for you that operator wants to suggest cheaper strage policy via sort of account info may exactly exists10:59
cbartzkota_: I understand. What still is missing is the information of the policy factor, which is not presented to the user. If given, the user could plan the capacity.10:59
kota_cbartz: it may be one way that we can solve my concern in the patch11:00
kota_to present the factor to users.11:00
cbartzkota_: Yes, currently, in a public-cloud scenario, the user does not care which storage policy to use, because it is transparent for him.11:00
kota_acoles: did you added me to a patch reviewer, for a while?11:01
kota_s/added/add/11:02
acoleskota_: yes on https://review.openstack.org/#/c/494172/ - I thought you might want to be aware of that change11:03
* kota_ got notification for patch review for... fixing EC frag duplication test!! nice!11:03
kota_acoles: ok, thanks for working that. it looks... medium volume to review :/11:04
cbartzkota_: btw, might you take a look an p 492935 ? Its related to the account spolicy, and its really a small patch. It fixes bug 1675776 .11:05
openstackbug 1675776 in OpenStack Object Storage (swift) "proxy.controllers.base.headers_to_account_info() should include per policy stats" [Wishlist,In progress] https://launchpad.net/bugs/1675776 - Assigned to Christopher Bartz (bartz)11:05
acoleskota_: the change lines are mostly moving/refactoring - the bug fix is to ensure that the final unique fragment is on a handoff not a primary11:05
kota_cbartz: not yet11:06
cbartzok11:06
*** manous has joined #openstack-swift11:09
kota_acoles: thanks the info that helps me to review11:12
*** tovin07_ has quit IRC11:16
*** tovin07 has quit IRC11:20
kota_acoles: done!11:32
acoleskota_: thank you11:32
*** abhitechie has joined #openstack-swift11:58
*** mat128 has joined #openstack-swift12:00
*** mvk has joined #openstack-swift12:04
*** mvk_ has quit IRC12:06
*** manous has quit IRC12:08
*** shuyingya has quit IRC12:25
*** shuyingya has joined #openstack-swift12:25
*** shuyingya has quit IRC12:29
*** abhitechie has quit IRC12:33
*** mat128 has quit IRC12:34
*** mat128 has joined #openstack-swift12:36
rcernintdasilva: ready in 10min?12:47
*** mvk has quit IRC12:48
*** catintheroof has joined #openstack-swift12:49
*** silor has joined #openstack-swift12:50
*** catintheroof has quit IRC12:50
*** catintheroof has joined #openstack-swift12:50
*** baojg has joined #openstack-swift12:55
openstackgerritChristopher Bartz proposed openstack/swift master: Account Quota Factor for Storage Policies.  https://review.openstack.org/48255813:09
*** lucasxu has joined #openstack-swift13:11
*** skudlik has quit IRC13:16
*** jogo_ has joined #openstack-swift13:21
jogo_hi there, please advise if there is any reason why i may use ceph instead of swift..13:22
jogo_am new to both storage technologies, however due to the goodies of distributed storage systems am confused on which option to use13:23
*** joeljwright has quit IRC13:28
*** joeljwright has joined #openstack-swift13:30
*** ChanServ sets mode: +v joeljwright13:30
*** ChanServ sets mode: +v jrichli13:32
tdasilvajogo_: of course you are asking a bunch of very biased folks ;) it's like asking at a ferrari dealership if you should buy a ferrari or a toyota ;) if what you need is object storage, we would obviously said use swift.13:48
tdasilvajogo_: but what is your use case? what are you trying to solve?13:48
*** mvk has joined #openstack-swift13:51
jogo_i have an old SAN running on former sun storagetek, because i need more storage, i want to move from SAN to software defined storage.. i have around 10 production vms and i need a scalable storage solution to be used to grow from 7tb to more13:51
jogo_easily13:51
*** skudlik has joined #openstack-swift13:53
tdasilvajogo_: so it sounds more like you need block storage??13:54
tdasilvafor your VMs?13:54
jogo_yes,13:54
tdasilvaah, then you want to go ceph13:55
tdasilvaswift is object storage only13:55
*** itlinux has joined #openstack-swift13:55
jogo_intresting thanks alot 031416113:55
*** MVenesio has joined #openstack-swift13:58
*** gyee has joined #openstack-swift14:02
*** vint_bra has joined #openstack-swift14:10
*** hseipp has joined #openstack-swift14:12
*** hseipp has quit IRC14:13
*** jogo_ has quit IRC14:16
*** mvk_ has joined #openstack-swift14:41
*** itlinux has quit IRC14:43
*** mvk has quit IRC14:44
*** mkrcmari__ has joined #openstack-swift14:45
*** itlinux has joined #openstack-swift14:46
*** hseipp has joined #openstack-swift14:46
*** mvk_ has quit IRC14:48
*** larivee has left #openstack-swift14:50
*** vinsh has quit IRC14:53
*** jeblair has joined #openstack-swift14:55
*** skudlik has quit IRC15:03
*** zaitcev has joined #openstack-swift15:18
*** ChanServ sets mode: +v zaitcev15:18
*** itlinux has quit IRC15:21
*** rcernin has quit IRC15:32
*** vint_bra has quit IRC15:37
*** cbartz has quit IRC15:38
*** vint_bra has joined #openstack-swift15:40
*** MVenesio has quit IRC15:41
*** tongli has joined #openstack-swift15:44
*** tongli has quit IRC15:44
*** d34dh0r53 has joined #openstack-swift16:01
*** patchbot has joined #openstack-swift16:02
notmynamegood morning16:03
notmynamehmm... seems patchbot died. I've restarted it16:03
*** rcernin has joined #openstack-swift16:08
timburkegood morning16:08
*** silor has quit IRC16:15
*** itlinux has joined #openstack-swift16:15
*** itlinux has quit IRC16:19
*** geaaru_ has joined #openstack-swift16:25
*** itlinux has joined #openstack-swift16:25
*** geaaru has quit IRC16:27
*** hseipp has quit IRC16:28
*** itlinux has quit IRC16:34
*** chsc has joined #openstack-swift16:35
*** MooingLemur has joined #openstack-swift16:37
*** itlinux has joined #openstack-swift16:53
*** abhitechie has joined #openstack-swift16:57
*** mkrcmari__ has quit IRC16:58
*** abhitechie has quit IRC16:59
*** abhitechie has joined #openstack-swift16:59
*** SkyRocknRoll has joined #openstack-swift17:08
*** shuyingya has joined #openstack-swift17:14
*** shuyingya has quit IRC17:18
*** MVenesio has joined #openstack-swift17:28
*** baojg has quit IRC17:39
*** joeljwright has quit IRC17:44
*** bkopilov has quit IRC17:44
*** abhitechie has quit IRC17:44
*** abhitechie has joined #openstack-swift17:44
*** itlinux has quit IRC17:53
*** itlinux has joined #openstack-swift17:58
*** oshritf has quit IRC18:00
*** oshritf has joined #openstack-swift18:07
*** oshritf has quit IRC18:08
*** xinli has joined #openstack-swift18:16
torgomaticI wrote a thing that lets you get all the stack traces out of a running Swift process: https://github.com/swiftstack/python-stack-xray18:19
torgomaticIt's handy when your daemon is behaving weirdly and logs + strace isn't enough18:19
notmynamefor those who haven't seen it yet, it's a really cool script. look at it, think, "oh ok". then look some more and think, "oh weird.... whoa! that's cool"18:20
*** xinli has quit IRC18:20
timburkei wonder how well it works when you have a mix of pthreads and greenthreads...18:35
*** abhitechie has quit IRC18:37
*** ChubYann has joined #openstack-swift18:39
tdasilvathe whole sequence of texts read kinda funny: "notmyname: hmm... seems patchbot died. I've restarted it. timburke: good morning"18:52
notmynameheh18:52
notmynametdasilva: you're learning the swiftstack secrets!!18:52
tdasilvalol18:52
tdasilvai think mattoliverau was right18:53
*** catintheroof has quit IRC18:54
*** vint_bra has quit IRC18:59
*** vint_bra has joined #openstack-swift19:02
*** itlinux has quit IRC19:04
*** itlinux has joined #openstack-swift19:05
*** silor has joined #openstack-swift19:12
*** itlinux has quit IRC19:12
*** SkyRocknRoll has quit IRC19:24
*** gkadam-afk has quit IRC19:35
*** itlinux has joined #openstack-swift19:47
*** silor has quit IRC19:58
*** MVenesio has quit IRC20:07
claygtdasilva: nice work on patch 472659!!!!20:14
patchbothttps://review.openstack.org/#/c/472659/ - swift - Allow to rebuild a fragment of an expired object20:14
claygrledisez: you too!20:14
tdasilvaclayg: putting a +2 after you and acoles have done so is much easier20:16
claygrofl20:16
*** rcernin has quit IRC20:16
tdasilvai think patch 364878 is also good to go, but thinking it might be better to wait until after release to merge it20:17
claygI plan to spend the next few years of my career just riding @acoles & @timburke's coattails (give @notmyname and @torgomatic a break from pulling my load)20:17
patchbothttps://review.openstack.org/#/c/364878/ - swift - Retrieve encryption root secret from Barbican20:17
claygoh yeah that'll be sweet20:18
notmynametdasilva: agreed. that should happen after the next release, please20:18
notmynamemathiasb: ^ FYI20:19
tdasilvanotmyname: ack20:19
claygtdasilva: it's pretty sweet tho - and you're talking about tagging like tomorrow right?20:19
notmynametdasilva: can you go ahead and +2 it with that comment? then next week you can +A it20:19
tdasilvaclayg: i think that was notmyname plan20:19
notmynameclayg: yeah. tomorrow ish20:19
torgomatictimburke: the stack xray thing works just fine with multiple greenthreads and pthreads; gc.get_objects() goes over everything the interpreter knows about, and it makes sure to get and release the GIL in the code it calls with gdb20:19
notmynameI'll do the authors/changelog patch tomorrow20:20
clayg@torgomatic so cool20:20
torgomatic@clayg between this, strace, and tcpdump, it's starting to get easier to figure out WTF a Swift daemon is doing when it goes off the rails :)20:21
clayg@torgomatic I wish like https://wiki.python.org/moin/DebuggingWithGdb just had that script?20:21
tdasilvaanybody read this: https://morepypy.blogspot.com/2017/08/lets-remove-global-interpreter-lock.html ? thoughts?20:21
torgomatic@clayg it wouldn't work unless you're using greenthreads though; it looks for things of type greenlet.greenlet and dumps their stacks20:22
*** vint_bra has quit IRC20:22
*** vint_bra has joined #openstack-swift20:23
*** vint_bra1 has joined #openstack-swift20:25
timburkeother fairly small things we might want to consider for the release: p 340250, p 451111, p 485381, p 476992, p 474801, p 47171220:26
patchbothttps://review.openstack.org/#/c/340250/ - swift - Add a jitter for check of time synchronization.20:26
patchbothttps://review.openstack.org/#/c/451111/ - swift - Add functests for disallowed COPYs into a versione...20:26
patchbothttps://review.openstack.org/#/c/485381/ - swift - functests: Only convert headers that should be int...20:26
patchbothttps://review.openstack.org/#/c/476992/ - swift - Make If-None-Match:* work properly with 0-byte PUTs20:26
patchbothttps://review.openstack.org/#/c/474801/ - swift - Ignore all auditor_status_*.json files in reconstr...20:26
patchbothttps://review.openstack.org/#/c/471712/ - swift - Improve domain_remap docs20:26
tdasilvatimburke: is that in order of priority?20:27
* timburke shrugs20:27
*** vint_bra has quit IRC20:27
timburkeabout half of them are tests-/docs-only20:28
timburkethe if-none-match guy is just a sneaky edge case i noticed a bit ago20:29
claygtdasilva: interesting #1 I hadn't realized that STM had failed #2 I hadn't realize that anyone thought gilectemy was a good idea (pypy seems to be suggesting to do fine grained locking and remove the gil but in pypy and without the ref count problem) #3 I mostly only want threads for blocking i/o, I don't *hate* using multiprocess to get the cores (better primitives for CSP would be welcome), I think that20:29
claygpushing i/o into threads managed by the eventloop (e.g. libuv/node.js/uvloop) is the most attractive answer for how I want to use threads20:29
claygtdasilva: thanks for sharing!20:30
tdasilvaclayg: and thanks for sharing your thoughts. I was curious to understand if this is something that could be beneficial to blocking i/o20:31
timburkei'm gonna look at patch 494172 -- already has a +2 from kota_ and everything!20:31
patchbothttps://review.openstack.org/#/c/494172/ - swift - Fix flakey EC frag duplication policy test20:31
claygI don't see how it makes what we tried to do with threadpool much better unfortunately :'(20:31
claygtho even with the stink on threads + python - I think now with servers per port - if we did thread per connection (on the storage servers) - we'd be ultimately happier than what we have now with blocking i/o in the event loop20:33
clayguvloop on pypy3 is the dream - but as redbo likes to point out - that's like saying you'll rewrite the object server in a language that hasn't been invented yet :D20:34
*** hoonetorg has quit IRC20:36
*** vint_bra1 has quit IRC20:43
*** vint_bra has joined #openstack-swift20:48
timburkethanks clayg! if you like, since you've already got that kinda loaded in your head, you might wanna check out https://review.openstack.org/#/c/468168/20:55
patchbotpatch 468168 - swift - Improve time_check output20:55
claygtimburke: that's not any better - that code is stupid - i hate recon-cli20:59
*** lucasxu has quit IRC21:01
*** tonyb has joined #openstack-swift21:05
*** mat128 has quit IRC21:05
*** shuyingya has joined #openstack-swift21:14
*** shuyingya has quit IRC21:19
*** geaaru_ has quit IRC22:08
*** itlinux has quit IRC22:23
*** itlinux has joined #openstack-swift22:25
*** vint_bra has quit IRC22:27
*** openstackgerrit has quit IRC22:33
*** itlinux has quit IRC22:35
*** openstackgerrit has joined #openstack-swift22:37
openstackgerritMerged openstack/swift master: Allow to rebuild a fragment of an expired object  https://review.openstack.org/47265922:37
openstackgerritMerged openstack/swift master: Fix flakey EC frag duplication policy test  https://review.openstack.org/49417222:39
timburkeyeah, merge all the stuff! there are two more in the gate right now :-)22:42
*** chsc has quit IRC23:38
timburkeoh, did we ever make a decision on https://review.openstack.org/#/c/476953/? it's kinda weird, 'cause https://review.openstack.org/#/c/476955/ is approved...23:44
patchbotpatch 476953 - swift - Make swift-dispersion-report importable23:44
patchbotpatch 476955 - swift - Add bin/swift-dispersion-report23:44

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!