Wednesday, 2017-07-05

*** NM has joined #openstack-swift00:30
*** tovin07_ has joined #openstack-swift00:45
kota_good morning01:03
* kota_ has been keeping cold a bit :/01:04
kota_i think U.S. people is still in holidays01:04
*** NM has quit IRC01:25
*** cshastri has joined #openstack-swift01:25
*** mat128 has quit IRC01:26
*** tovin07__ has joined #openstack-swift01:37
*** tovin07_ has quit IRC01:39
*** bkopilov has quit IRC01:59
*** ChubYann has quit IRC02:06
*** openstackgerrit has joined #openstack-swift02:11
openstackgerritliuyamin proposed openstack/swift master: Add license in swift code file  https://review.openstack.org/48032702:11
*** ChubYann has joined #openstack-swift02:19
*** abhitechie has joined #openstack-swift02:57
*** happyyi008 has joined #openstack-swift03:03
*** abhitechie has quit IRC03:05
*** abhitechie has joined #openstack-swift03:05
*** bkopilov has joined #openstack-swift03:13
*** SkyRocknRoll has joined #openstack-swift03:24
*** gkadam has joined #openstack-swift03:24
*** abhitechie has quit IRC03:25
*** gkadam has quit IRC03:26
*** gkadam has joined #openstack-swift03:27
*** sanchitmalhotra has joined #openstack-swift03:33
*** kei_yama has quit IRC03:42
*** links has joined #openstack-swift03:44
*** happyyi008 has quit IRC03:45
*** kei_yama has joined #openstack-swift04:02
*** caiobrentano has quit IRC04:30
*** psachin has joined #openstack-swift04:32
*** mtreinish has quit IRC04:44
*** mtreinish has joined #openstack-swift04:51
*** ChubYann has quit IRC05:00
*** ukaynar has quit IRC05:34
*** ukaynar has joined #openstack-swift05:34
*** ukaynar_ has joined #openstack-swift05:43
*** ukaynar has quit IRC05:43
*** cschwede has joined #openstack-swift06:25
*** ChanServ sets mode: +v cschwede06:25
*** ukaynar_ has quit IRC07:02
*** tesseract has joined #openstack-swift07:02
*** ukaynar has joined #openstack-swift07:03
*** ukaynar has quit IRC07:07
*** rcernin has joined #openstack-swift07:19
*** oshritf has joined #openstack-swift07:36
*** d0ugal has joined #openstack-swift07:46
*** d0ugal has quit IRC07:46
*** d0ugal has joined #openstack-swift07:46
acolesgood morning07:52
acoleskota_: get well soon!07:55
openstackgerritMerged openstack/swift master: Add license in swift code file  https://review.openstack.org/48032707:57
kota_acoles: thx!08:18
*** cbartz has joined #openstack-swift08:29
*** links has quit IRC08:37
*** dancn has joined #openstack-swift08:39
*** links has joined #openstack-swift08:53
*** kei_yama has quit IRC08:56
dancncschwede: feel free to ping me here for https://review.openstack.org/#/c/473862/09:07
patchbotpatch 473862 - swift - Add full working example of sharing a container wi...09:07
*** links has quit IRC09:49
*** mat128 has joined #openstack-swift09:53
*** mat128 has quit IRC09:54
*** mat128 has joined #openstack-swift09:58
*** links has joined #openstack-swift10:03
*** oshritf has quit IRC10:16
*** oshritf has joined #openstack-swift10:23
*** cshastri has quit IRC10:26
*** tovin07__ has quit IRC10:28
*** oshritf has quit IRC10:33
*** oshritf has joined #openstack-swift10:41
*** NM has joined #openstack-swift10:42
*** NM has quit IRC10:48
*** bkopilov has quit IRC10:59
*** caiobrentano has joined #openstack-swift11:16
*** ianychoi has quit IRC11:28
*** oshritf has quit IRC11:43
*** oshritf has joined #openstack-swift11:48
*** dancn has left #openstack-swift12:03
*** ouchkernel has joined #openstack-swift12:07
*** ouchkern_ has joined #openstack-swift12:13
*** ouchkernel has quit IRC12:16
*** caiobrentano has quit IRC12:24
*** bkopilov has joined #openstack-swift12:40
*** lucasxu has joined #openstack-swift13:00
*** caiobrentano has joined #openstack-swift13:06
*** catintheroof has joined #openstack-swift13:18
openstackgerritOpenStack Proposal Bot proposed openstack/swift master: Updated from global requirements  https://review.openstack.org/8873613:22
*** links has quit IRC13:24
*** joelinux73 has joined #openstack-swift13:29
*** joelinux_ has joined #openstack-swift13:31
tdasilvagood morning13:31
*** joelinux_ has quit IRC13:32
*** gkadam has quit IRC13:33
*** joelinux73 has quit IRC13:33
*** joelinux73 has joined #openstack-swift13:33
*** joelinux_ has joined #openstack-swift13:34
*** joelinux_ has quit IRC13:35
*** joelinux73 has quit IRC13:35
*** joelinux73 has joined #openstack-swift13:35
*** joelinux73 has quit IRC13:36
*** joelinux73 has joined #openstack-swift13:36
*** klamath has joined #openstack-swift13:37
*** vint_bra has quit IRC13:37
*** klamath has quit IRC13:38
*** klamath has joined #openstack-swift13:38
acolestdasilva: o/13:41
*** zhurong has joined #openstack-swift13:44
*** NM has joined #openstack-swift13:52
tdasilvaacoles: o/13:58
*** zhurong has quit IRC14:09
*** chlong_ has joined #openstack-swift14:31
openstackgerritMerged openstack/swift master: Add support to increase object ring partition power  https://review.openstack.org/33729714:40
*** vint_bra has joined #openstack-swift14:57
*** lucasxu has quit IRC15:00
openstackgerritOpenStack Proposal Bot proposed openstack/swift master: Updated from global requirements  https://review.openstack.org/8873615:01
*** cebruns__ is now known as cebruns15:03
*** bkopilov has quit IRC15:14
*** rcernin has quit IRC15:21
*** chlong has joined #openstack-swift15:23
*** chlong has quit IRC15:26
*** lucasxu has joined #openstack-swift15:35
*** gyee has joined #openstack-swift15:36
*** psachin has quit IRC15:37
*** bkopilov has joined #openstack-swift15:40
*** chlong_ has quit IRC15:58
*** chlong_ has joined #openstack-swift16:14
*** JimCheung has joined #openstack-swift16:24
*** lucasxu has quit IRC16:27
*** mat128 has quit IRC16:29
*** abhitechie has joined #openstack-swift16:31
*** mat128 has joined #openstack-swift16:33
*** cbartz has quit IRC16:36
openstackgerritGang Xu proposed openstack/swift master: Fix html_last_updated_fmt for Python3  https://review.openstack.org/48067016:40
timburkegood morning16:41
*** cebruns has quit IRC16:50
acolestimburke: welcome back16:54
*** cebruns has joined #openstack-swift16:54
*** itlinux has joined #openstack-swift16:55
*** zaitcev has joined #openstack-swift17:06
*** ChanServ sets mode: +v zaitcev17:06
*** chsc has joined #openstack-swift17:19
*** chsc has joined #openstack-swift17:19
caiobrentanohey tdasilva, do you know any RPM repository where I could find liberasurecode package?17:19
tdasilvacaiobrentano: http://mirror.centos.org/centos/7/cloud/x86_64/17:20
tdasilvacaiobrentano: that's for centos17:20
tdasilvacaiobrentano: so, for newton it will be here: http://mirror.centos.org/centos/7/cloud/x86_64/openstack-newton/common/17:21
zaitcevone of fedora mirros has it too17:21
*** chlong_ has quit IRC17:21
*** abhitechie has quit IRC17:22
tdasilvazaitcev: btw, we released 1.5.0 last week, I think it is worth getting into pike ??17:23
tdasilvatimburke: ^^^ there was some good discussions last week here around one of your patches, worth  reading up on it17:23
tdasilvatimburke: I think we also left some comments on the patch itself17:24
zaitcevI pretty much gave up on trying to do that. The best we can do at this point is to file a bug and have jjoyce do it.17:24
timburketdasilva: yeah, thanks. it's on my list of things to circle back on this week :-)17:24
tdasilvatimburke: this: https://review.openstack.org/#/c/459023/17:24
patchbotpatch 459023 - liberasurecode - Consistently use zlib for crc3217:24
caiobrentanotdasilva cool! I was looking for el6... but I'll check this one17:24
tdasilvazaitcev: ack, will do that17:24
caiobrentanotdasilva thanks!17:25
tdasilvacaiobrentano: oh sorry, i'm not sure we build el6 anymore17:25
tdasilvabut i could be wrong17:25
zaitcevwait, no. That zlib patch was very suspicious17:25
zaitcevoh, I actually commented on it. Well, carry on17:25
tdasilvaheheh17:25
zaitcevtimburke, I have a more important question for you: do you like that cschwede's part++ thing? I think I like it, but you guys caught me not reading it closely enough couple of times, last time with an unused argument17:26
zaitcevOooh17:27
zaitcevit's merged17:27
zaitcevyay17:27
zaitcevokay, I can eat cookies and drink milk for the rest of the day today then17:27
timburkeyay cookies and milk!17:28
zaitcevSo, what I meant to say about 459023 is, it looks to me that zlib uses a different polynomial that we appear to be using, therefore the CRC of metadatas that are already stored will mismatch after upgrade.17:29
zaitcevBut I totally can be wrong about it.17:29
*** chlong_ has joined #openstack-swift17:33
tdasilvado we have team meeting today?17:36
tdasilvacommunity meeting, whatever it's calle17:36
timburkeso the #define at https://github.com/openstack/liberasurecode/blob/1.5.0/src/utils/chksum/crc32.c#L115-L116 *does* match the SSE4-optimized CRC-32C -- but we never actually use it anywhere!17:38
timburkemeanwhile, the table at https://github.com/openstack/liberasurecode/blob/1.5.0/src/utils/chksum/crc32.c#L69-L113 *almost* matches the one at https://github.com/madler/zlib/blob/v1.2.11/crc32.h#L8-L5917:40
*** ChubYann has joined #openstack-swift17:40
timburke...except we're using signed ints, which shift differently!17:40
timburkeso we're doing some completely different, somewhat-crazy CRC-like thing whose properties are not well understood17:41
timburkeexcept for all those times where we have zlib already loaded so *none of it mattered at all* and it's all dead code17:42
tdasilvatimburke: yeah, but running that little test script you mentioned seemed to point to an unexpected result17:44
tdasilvawhere zlib is not be loaded as you expected??17:45
timburkeyeah, i think i need to update it to account for frags that don't have a crc stored17:45
notmynamegood morning17:46
timburkein which case it's all *definitely* moot; the check is basically bypassed17:46
notmynametdasilva: yes, there is a meeting today at 210017:46
tdasilvanotmyname: cool! thanks17:47
notmynameI've been out, so only just now catching up to things :-)17:47
tdasilvanotmyname: too much bbq? :D17:48
*** gkadam has joined #openstack-swift17:48
notmynameI wish (or not. CA bbq is not good)17:48
notmynamewe had a great long weekend, though. watching fireworks. time with family and friends17:48
tdasilvanotmyname: :)17:51
notmynamewhoa! part power increase landed!17:51
timburkeyeah it did!17:52
openstackgerritClay Gerrard proposed openstack/swift master: Replace replication_one_per_device by custom count  https://review.openstack.org/39078118:00
claygcschwede: kong: mattoliverau: is anyone going to be around for the meeting today to explain to me how patch 470158 works now?18:03
patchbothttps://review.openstack.org/#/c/470158/ - swift - Write-affinity aware object deletion18:03
*** tonanhngo has joined #openstack-swift18:08
openstackgerritSamuel Merritt proposed openstack/swift master: Unify Putter and MIMEPutter into a single class.  https://review.openstack.org/34052618:09
tdasilvatorgomatic: \o/ !!!18:10
openstackgerritSamuel Merritt proposed openstack/swift master: Save ring builder if dispersion changes  https://review.openstack.org/47901218:10
torgomatictdasilva: I finally dug myself out of the other pile of work I had. :) Only took a year...18:12
tdasilvatorgomatic: welcome back!18:12
tdasilvayou were missed18:12
torgomaticthanks :)18:12
zaitcevand he starts out by putting sticks into the wheels of 427911, PUT+POST :-)18:13
zaitcevI need the extra class to perform easy switching!18:13
claygzaitcev: I think that was just an old patch he was cleaning up - and also I'm pretty sure he's interested in what to do about PUT+POST18:14
torgomaticI should have waited a couple more weeks for that patch; 340526 was about to have its first birthday18:15
zaitcevoh, I wasn't aware, sorry18:16
zaitcevMaybe I can build on top of it...18:16
*** gkadam has quit IRC18:19
*** cebruns has quit IRC18:21
*** cebruns has joined #openstack-swift18:26
*** rcernin has joined #openstack-swift18:47
*** chlong_ has quit IRC19:14
*** chlong_ has joined #openstack-swift19:30
zaitcevhttps://twitter.com/kurogane_s/status/88249618351133491219:30
claygi have no idea what is going on19:33
*** karenc has quit IRC19:39
*** joeljwright has joined #openstack-swift19:48
*** ChanServ sets mode: +v joeljwright19:48
*** mat128 has quit IRC19:50
*** ChanServ changes topic to "Queens PTG planning https://etherpad.openstack.org/p/swift-ptg-queens | Logs: http://eavesdrop.openstack.org/irclogs/%23openstack-swift/ | Meetings: https://wiki.openstack.org/wiki/Meetings/Swift | Review Dashboard: http://not.mn/reviews.html"19:53
*** chlong_ has quit IRC19:55
*** tesseract has quit IRC19:56
*** NM has quit IRC19:57
*** NM has joined #openstack-swift19:59
*** rcernin has quit IRC20:05
*** ouchkern_ has quit IRC20:12
*** joeljwright has quit IRC20:13
*** rcernin has joined #openstack-swift20:14
*** NM has quit IRC20:23
timburketdasilva: got an example frag?20:23
tdasilvatimburke: uh, let me see if I can find it20:24
*** caiobrentano has quit IRC20:26
*** caiobrentano has joined #openstack-swift20:26
*** NM has joined #openstack-swift20:29
*** caiobrentano_ has joined #openstack-swift20:29
*** caiobrentano has quit IRC20:31
*** joeljwright has joined #openstack-swift20:31
*** ChanServ sets mode: +v joeljwright20:31
*** chlong_ has joined #openstack-swift20:44
*** NM has quit IRC20:51
clayg@timburke so you said my frag didn't have the checksum info - which makes sense enough to me looking at the parsed frag metadata and the hexdump - but like... is this ec frag hulla old or is the checksum business sorta recent-ish?20:52
notmynameyikes. the openstack-manuals project repo is 767MB. 730MB of that is the .git directory20:52
claygoic, pre-1.2.020:53
timburkeclayg: the frag was written with libec 1.0.9 -- you can see it in bytes 0x3f-0x4220:53
claygnowhey!20:54
timburkelooks like that was put out in September 201520:54
claygso this is one I wrote this morning into the Sac-Eng cluster20:54
clayg[root@sac-eng-paco02 ~]# ./check_crc.py /srv/node/d5/objects-1/58879/5de/e5ff296618741e472a3a66c9689725de/1498869266.60322#3#d.data20:54
clayg/srv/node/d5/objects-1/58879/5de/e5ff296618741e472a3a66c9689725de/1498869266.60322#3#d.data: ERROR: Checksum mismatch: 1873f8ec != b945ee2220:54
clayg... with the new script20:55
kota_morning20:57
clayg[root@sac-eng-paco02 ~]# LD_PRELOAD=/opt/ss/lib64/liberasurecode.so.1.4.0 ./check_crc.py /srv/node/d5/objects-1/58879/5de/e5ff296618741e472a3a66c9689725de/1498869266.60322#3#d.data20:58
clayg/srv/node/d5/objects-1/58879/5de/e5ff296618741e472a3a66c9689725de/1498869266.60322#3#d.data: ok20:58
mattoliverauMorning20:58
claygoh... I think that's bad maybe?  or maybe I don't understand yet...20:58
notmynamemeeting time in #openstack-meeting21:00
*** m_kazuhiro has joined #openstack-swift21:00
*** NM has joined #openstack-swift21:04
timburkeclayg: well 💩. guess I gotta support both :-(21:07
*** NM has quit IRC21:07
kota_clayg, timburke: if older swift had a chance to use liberasurecode's crc in existing fragments, i vote we cannot revert liberasurecode crc soon. at least, we need keep it and long term deprecating for backword compat, right?21:10
tdasilvaif ever, right? i mean, data lives for a long time.............21:11
kota_if ever, yes tdasilva21:11
timburkeyup. gotta handle both, but still wanna start using zlib's implementation for any new frags (because, again, that's the one that actually has some degree of mathematical rigor behind it)21:12
tdasilvatimburke: sounds good21:12
kota_otherwise, the newer liberasurecode may quarantine the correct frags...21:12
kota_timburke: yup, +1 for using zlib for newr frags21:13
*** chlong_ has quit IRC21:18
*** NM has joined #openstack-swift21:18
*** NM has quit IRC21:22
*** caiobrentano_ has quit IRC21:34
*** caiobrentano has joined #openstack-swift21:35
*** caiobrentano_ has joined #openstack-swift21:37
*** caiobrentano has quit IRC21:39
notmynameFYI acoles lost power/internet21:52
acolessorry I dropped out of meeting then, power gone, probably something to do with brexit, hopefully will be fixed while I sleep21:54
mattoliverauacoles: lol21:55
acoleshmmm, phone battery almost dead too. good night.21:55
kota_hmm... it looks, perheps, 5>=parities could cover almost condition on https://github.com/01org/isa-l/issues/10#issuecomment-310944022 ?21:55
tdasilvaacoles: good night21:56
mattoliverauk, baby has had me up since 4am, so I'm going to find food and then take a nap ;) (benefits of unemployement)21:56
kota_acoles: good night21:56
tdasilvataking off too, will try to checkin back later21:56
* kota_ needs clean heads in around daytime. now it's still a bit early to think deeeply21:56
*** joeljwright has quit IRC21:57
kota_weird21:58
*** NM has joined #openstack-swift21:58
*** m_kazuhiro has quit IRC21:58
*** NM has quit IRC21:58
kota_the issue reported only k <= 21, m-k = 4 is safe so it seems not to relate to ceph's condition https://github.com/ceph/ceph/blob/master/src/erasure-code/isa/ErasureCodeIsa.cc#L38121:59
kota_k <= 21, m-k = 4 means sort of 21, 25 for k, m22:00
kota_not 21, 422:00
*** NM has joined #openstack-swift22:01
kota_let's try later...22:01
kota_(but at least today, daytime922:01
*** NM has quit IRC22:01
kota_)22:01
* kota_ is findin breakfast22:10
*** klamath has quit IRC22:14
*** tonanhngo has quit IRC22:21
*** tonanhngo has joined #openstack-swift22:26
*** tonanhngo has quit IRC22:31
notmynamejungleboyj: seems that swift docs migration might be pretty simple22:31
jungleboyjnotmyname: :-)  Color me jealous.22:32
notmynameit's only the admin guide that has content in openstack-manuals that needs to move to the swift repo22:32
notmynamewe moved the install guide a long time ago, so that's already in swift22:32
notmynameand we already have the api guide in swift too22:33
notmynamethere's not a lot of content in manuals that needs to move. there's more stuff that maybe *should* have been there, and there's more docs (ops guides etc) that are already in the swift repo22:34
jungleboyjnotmyname:  Ok.  That is good.  Part of the problem is that we didn't have the right sphinx options set and now that I have enabled them I am finding all kinds of broken issues?22:42
notmynamejungleboyj: like "error: error in setup.cfg: command 'build_sphinx' has no such option 'warning_is_error'"22:44
jungleboyjNo, Doug noted I should have warning_is_error=1 set and now I am finding that we had a lot of warning being ignored.22:45
*** rcernin has quit IRC22:45
*** itlinux has quit IRC22:50
*** caiobrentano_ has quit IRC22:52
*** caiobrentano has joined #openstack-swift22:52
*** caiobrentano has quit IRC22:57
*** vint_bra has quit IRC22:57
*** JimCheung has quit IRC22:59
*** JimCheung has joined #openstack-swift23:00
*** ^andrea^ has quit IRC23:18
thurloatcould a rebalance cause objects to 404 until they get moved to the proper location? or should they be resolved before it completes?23:20
torgomaticjust one rebalance shouldn't cause problems, but if you rebalance over and over before the replicators have finished moving stuff, then you might get 404s23:24
*** kei_yama has joined #openstack-swift23:28
*** chsc has quit IRC23:35
thurloattorgomatic thanks, I did indeed to that.23:42
thurloathad some disk space issues that needed immediate addressing and did a few rebalances in quick succession23:42
*** JimCheung has quit IRC23:43
thurloatbut the replicators should catch up eventually?23:43
*** JimCheung has joined #openstack-swift23:43
torgomaticthurloat: yes, they'll sort things out eventually.23:46
thurloatfigured as much, thanks for the confidence though torgomatic23:46
torgomaticnp23:47
*** hoonetorg has quit IRC23:47
*** catintheroof has quit IRC23:47

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