Thursday, 2017-10-12

*** tovin07_ has joined #openstack-swift00:46
*** klrmn has quit IRC01:12
*** nikivi has quit IRC01:27
*** nikivi has joined #openstack-swift01:33
*** psachin has joined #openstack-swift01:49
*** Shatadru|Gone is now known as Shatadru|coffee|02:16
* kota_ is back02:52
*** klrmn has joined #openstack-swift02:55
*** mtreinish has quit IRC03:02
*** mtreinish has joined #openstack-swift03:07
*** wes_dillingham has quit IRC03:11
*** wes_dillingham has joined #openstack-swift03:12
*** wes_dillingham has quit IRC03:12
*** wes_dillingham has joined #openstack-swift03:13
*** wes_dillingham has quit IRC03:13
*** wes_dillingham has joined #openstack-swift03:13
*** wes_dillingham has quit IRC03:14
*** wes_dillingham has joined #openstack-swift03:14
*** wes_dillingham has quit IRC03:14
*** wes_dillingham has joined #openstack-swift03:15
*** wes_dillingham has quit IRC03:15
*** wes_dillingham has joined #openstack-swift03:16
*** wes_dillingham has quit IRC03:16
*** wes_dillingham has joined #openstack-swift03:17
*** wes_dillingham has quit IRC03:17
*** wes_dillingham has joined #openstack-swift03:17
*** wes_dillingham has quit IRC03:18
*** wes_dillingham has joined #openstack-swift03:18
*** wes_dillingham has quit IRC03:18
*** wes_dillingham has joined #openstack-swift03:20
*** wes_dillingham has quit IRC03:20
*** wes_dillingham has joined #openstack-swift03:22
*** wes_dillingham has quit IRC03:22
*** wes_dillingham has joined #openstack-swift03:22
*** wes_dillingham has quit IRC03:22
*** wes_dillingham has joined #openstack-swift03:23
*** wes_dillingham has quit IRC03:23
*** wes_dillingham has joined #openstack-swift03:24
*** wes_dillingham has quit IRC03:24
*** links has joined #openstack-swift03:24
*** wes_dillingham has joined #openstack-swift03:25
*** wes_dillingham has quit IRC03:25
*** gkadam has joined #openstack-swift04:09
*** SkyRocknRoll has joined #openstack-swift04:21
*** klrmn has quit IRC04:30
*** hoonetorg has quit IRC04:37
*** hoonetorg has joined #openstack-swift04:50
*** jarbod has quit IRC04:57
*** jarbod has joined #openstack-swift04:58
*** ChubYann has quit IRC05:28
*** HCLTech-SSW has joined #openstack-swift05:32
HCLTech-SSW<Clay Gerrard>: Hi05:34
*** gkadam_ has joined #openstack-swift05:38
*** gkadam has quit IRC05:40
*** gkadam__ has joined #openstack-swift05:40
*** gkadam__ has quit IRC05:42
*** gkadam_ has quit IRC05:43
*** gkadam has joined #openstack-swift05:43
*** Shatadru|coffee| is now known as Shatadru|brb05:45
*** cshastri has joined #openstack-swift05:45
kota_timburke: hmm... odd. I couldn't reproduce what I did, on the patch 345739.s3.amazonaws.com also work the patch expected.05:49
patchbothttps://review.openstack.org/#/c/345739/ - swift3 - Add support for more characters in header keys05:49
kota_maybe i was dreaming... ?05:49
kota_timburke: gaah! I noticed my mistake. The hint is in my comment at the gerrit. I was using both x-amz-meta-foo and x-amz-meta_foo but actually s3 requiers 'x-amz-meta-' prefix (not that the dash after "meta")06:00
kota_and only one correct header (i.e. x-amz-meta-foo) was remained in the object metadata. whooa, sorry :(06:01
*** Shatadru|brb is now known as Shatadru06:35
*** rcernin has joined #openstack-swift06:35
*** HCLTech-SSW has quit IRC06:37
*** gkadam has quit IRC06:40
openstackgerritMerged openstack/swift feature/deep: Replicate deleted shard ranges  https://review.openstack.org/51128206:41
*** gkadam has joined #openstack-swift06:42
*** Dinesh_Bhor has quit IRC07:01
*** Dinesh_Bhor has joined #openstack-swift07:07
*** hseipp has joined #openstack-swift07:08
*** pcaruana has joined #openstack-swift07:17
*** gkadam is now known as gkadam-afk07:24
*** Shatadru is now known as Shatadru|food07:27
*** tesseract has joined #openstack-swift07:32
*** oshritf has joined #openstack-swift07:40
*** oshritf has quit IRC07:48
*** d0ugal has joined #openstack-swift08:00
openstackgerritOpenStack Proposal Bot proposed openstack/swift master: Imported Translations from Zanata  https://review.openstack.org/51000008:12
openstackgerritTovin Seven proposed openstack/swift master: OSprofiler in OpenStack Swift  https://review.openstack.org/46831608:15
acolesgood morning08:20
*** gkadam-afk has quit IRC08:28
*** gkadam has joined #openstack-swift08:31
*** kei_yama has quit IRC08:38
openstackgerritKota Tsuyuzaki proposed openstack/swift master: WIP: Move swift-drive-audit code to swift/cli/drive_audit  https://review.openstack.org/51120508:40
*** spectr has quit IRC08:49
*** Shatadru|food is now known as Shatadru08:53
mattoliverauacoles: morning09:11
acolesmattoliverau: o/09:11
openstackgerritAlistair Coles proposed openstack/swift feature/deep: Add helper methods to assert shard range equality  https://review.openstack.org/51143709:18
*** spectr has joined #openstack-swift09:19
tdasilvagood morning09:19
acolestdasilva: ? you're early09:19
tdasilvajust a bit :)09:20
tdasilvawanted to catch up with kota_ and m_kazuhiro09:20
tdasilvareading scrollback09:20
*** gkadam_ has joined #openstack-swift09:36
*** mvk has quit IRC09:37
*** gkadam has quit IRC09:38
openstackgerritAlistair Coles proposed openstack/swift feature/deep: Merge put_shard_range and delete_shard_range to one update method  https://review.openstack.org/51144609:46
*** tovin07_ has quit IRC10:00
*** Shatadru is now known as Shatadru|Gone10:06
*** spectr has quit IRC10:12
*** m_kazuhiro has joined #openstack-swift10:14
m_kazuhirogood morning.10:14
kota_tdasilva: hello10:15
kota_tdasilva: sorry, I have to leave here now, m_kazuhiro will be able to continue to discuss something10:15
tdasilvahello10:15
tdasilvano problem10:15
kota_I'll also catch up with reading the irc log ;)10:16
tdasilvaok :)10:16
kota_see you10:16
tdasilvam_kazuhiro: just wanted to catch up on the issue #210:16
m_kazuhirotdasilva: On my etherpad page?10:16
tdasilvayeah10:16
tdasilvaso I'm not against adding the automatic content type, but was just trying to understand the use case/motivation for doing so10:18
m_kazuhiroIn my understanding, there are no use case for auto insertion of content-type.10:19
m_kazuhiroThis topic is come from ``application/symlink`` is recommended in comment of symlink middleware.10:20
tdasilvaah ok...so we might as well add it if users doesn't send a content-type?10:22
m_kazuhiroSo my opinion is that if there is recommended content-type, the content-type should be inserted automatically. But I don't have motivation for recommended content-type.10:22
m_kazuhiroYes auto insertion is topic when user doesn't send content-type.10:23
m_kazuhiroBefore start discussion, I want to ask one question.10:24
m_kazuhiroAre there any reason why you wrote about "recommended content-type"? I think you wrote the comment.10:25
*** spectr has joined #openstack-swift10:25
m_kazuhiroComment is at https://review.openstack.org/#/c/232162/59/swift/common/middleware/symlink.py@3410:25
patchbotpatch 232162 - swift - Symlink implementation.10:25
tdasilvam_kazuhiro: i'm trying to remember the reason for that TBH. I think it is basically another way to distinguish the object. Can be useful for the app devs.10:27
tdasilvam_kazuhiro: ok, I started on a patch based on kota's comments and will try to send an update today10:28
m_kazuhirotdasilva: Will you update symlink patch? I think you are busy for swift3 work. So if you say OK, I will update it.10:31
tdasilvam_kazuhiro: i got started on the symlink patch yesterday, so I think I should be able to send a quick update...10:32
m_kazuhirotdasilva: Sounds great. OK, I will review updated symlink patch after your update.10:34
*** Shatadru|Gone has quit IRC10:34
tdasilvam_kazuhiro: ok.10:36
tdasilvai'm going to go get ready to drive to the office, thanks for sticking around :)10:36
m_kazuhirotdasilva: You're welcome. I'm looking forward to patch update :)10:37
*** m_kazuhiro has quit IRC10:52
*** geaaru has joined #openstack-swift11:08
*** mvk has joined #openstack-swift11:08
*** oshritf has joined #openstack-swift11:14
openstackgerritMerged openstack/swift feature/deep: Shard ranges don't need storage policies  https://review.openstack.org/51131011:16
*** cbartz has joined #openstack-swift11:57
*** cbartz has quit IRC12:01
*** Shatadru|Gone has joined #openstack-swift12:14
*** wes_dillingham has joined #openstack-swift12:14
*** cbartz has joined #openstack-swift12:18
*** Shatadru|Gone has quit IRC12:20
*** lifeless has quit IRC12:24
*** lifeless has joined #openstack-swift12:31
*** Shatadru has joined #openstack-swift12:40
*** links has quit IRC12:41
*** gkadam_ has quit IRC12:46
*** catintheroof has joined #openstack-swift13:06
*** catintheroof has quit IRC13:06
*** catintheroof has joined #openstack-swift13:06
*** mat128 has joined #openstack-swift13:13
*** catintheroof has quit IRC13:13
*** catintheroof has joined #openstack-swift13:14
*** SkyRocknRoll has quit IRC13:21
*** oshritf has quit IRC13:24
*** chlong has joined #openstack-swift13:25
*** cshastri has quit IRC13:27
openstackgerritAlistair Coles proposed openstack/swift feature/deep: clean up _one_shard_cycle state machine  https://review.openstack.org/51148213:38
*** cshastri has joined #openstack-swift13:43
*** Shatadru has quit IRC13:51
*** itlinux has joined #openstack-swift13:58
*** oshritf has joined #openstack-swift14:10
*** itlinux has quit IRC14:12
*** chlong has quit IRC14:13
*** chlong has joined #openstack-swift14:18
*** itlinux has joined #openstack-swift14:24
*** cshastri has quit IRC14:25
*** psachin has quit IRC14:28
*** Shatadru has joined #openstack-swift14:30
*** itlinux has quit IRC14:32
*** itlinux has joined #openstack-swift14:36
*** links has joined #openstack-swift14:39
*** rcernin has quit IRC14:55
-openstackstatus- NOTICE: Job log uploads are failing due to lack of inodes. Jobs also fail due to mismatches in gnutls packages. Workarounds for both in progress with proper fixes to follow.14:55
*** ChanServ changes topic to "Job log uploads are failing due to lack of inodes. Jobs also fail due to mismatches in gnutls packages. Workarounds for both in progress with proper fixes to follow."14:56
*** chlong has quit IRC15:02
*** mjturek has joined #openstack-swift15:06
mjturekhey - I'm running tempest against a non-devstack environment and I was wondering if the tempest tests expect a certain auth type to be used? I'm getting a lot of "Forbidden" responses ie - http://paste.openstack.org/show/623463/15:11
*** chlong has joined #openstack-swift15:15
*** oshritf has quit IRC15:16
*** itlinux has quit IRC15:17
*** klrmn has joined #openstack-swift15:30
openstackgerritAlistair Coles proposed openstack/swift feature/deep: make root_account, root_container properties of broker  https://review.openstack.org/51150815:38
timburkegood morning15:41
*** itlinux has joined #openstack-swift15:42
timburkekota_: no worries on the metadata patch. you just had me doubting myself :-)15:43
*** links has quit IRC15:47
timburkemjturek: i'd expect that tempest wants some flavor of keystone auth, but i don't really know for sure. assuming that you've already got that covered, my best guess for where to start looking would be to compare the roles for your test user and the operator_roles in the proxy-server's [filter:keystoneauth] section -- might be worth looking at https://github.com/openstack/swift/blob/2.15.1/etc/proxy-server.conf-sample#L410 too15:55
notmynamegood morning16:00
*** links has joined #openstack-swift16:00
*** chlong has quit IRC16:03
mjturekawesoome, thanks timburke!16:09
*** cbartz has quit IRC16:13
*** hseipp has quit IRC16:27
*** pdardeau has joined #openstack-swift16:30
*** itlinux has quit IRC16:32
*** gyee has joined #openstack-swift16:41
*** gyee has quit IRC16:44
*** geaaru has quit IRC16:49
*** mvk has quit IRC17:07
*** tesseract has quit IRC17:11
*** not_me has quit IRC17:14
*** silor has joined #openstack-swift17:18
*** oshritf has joined #openstack-swift17:19
*** Shatadru has quit IRC17:25
*** links has quit IRC17:27
*** silor has quit IRC17:38
openstackgerritTim Burke proposed openstack/swift feature/deep: Add helper methods to assert shard range equality  https://review.openstack.org/51143717:58
*** shaner has quit IRC18:00
claygohai!?18:03
*** shaner has joined #openstack-swift18:06
openstackgerritAlistair Coles proposed openstack/swift feature/deep: actually use the assertShardRangesNotEqual helper method  https://review.openstack.org/51154418:07
acolestimburke: thanks for all your careful review!18:08
timburkenp18:08
acolestimburke: mattoliverau I've noticed that the sharder probe tests seem to rely on timely calls to container list in order to pass... I think maybe because calls to the container server API (including shard range updates) land in a X.db.pending file even when the container has sharded to X_shard.db ! And it takes an API call to have that pending file emptied. https://github.com/openstack/swift/blob/b48a8f634d2539d7eb0ac51b218:18
acolesed2afc6ffc79c29/swift/container/server.py#L142-L15818:18
acolesIn the sharder, whenever we do get_info() on a sharded broker hoping it will persist all pending updates, it looks for *X_shard.db.pending*18:20
timburkeeep18:20
acolestimburke: I've been going nuts trying to debug the shrinking probe test, I *hope* this might be hallelujah moment18:21
acolesbut I have to stop for today! so, thoughts...should we always use one pending file with one name? or is there a reason to switch to X_shard.db.pending once sharded? I am thinking one file, same name, always, makes sense??18:23
acolesmattoliverau: timburke: ^^ any opinions?18:23
acolesgood night18:23
*** vint_bra has joined #openstack-swift18:29
timburkegoo night acoles! single name seems pretty sane to me, but i haven't thought through it much yet18:31
openstackgerritMerged openstack/swift feature/deep: Merge put_shard_range and delete_shard_range to one update method  https://review.openstack.org/51144618:34
mattoliverauacoles: oops, yeah should be 1 file. Because there should always only be 1 db to update in all states.18:37
*** joeljwright has joined #openstack-swift18:38
*** joeljwright has quit IRC18:39
*** hseipp has joined #openstack-swift18:40
*** ChubYann has joined #openstack-swift18:41
*** joeljwright has joined #openstack-swift18:41
*** ChanServ sets mode: +v joeljwright18:41
openstackgerritSamuel Merritt proposed openstack/swift master: Use "poll" or "selects" Eventlet hub for all Swift daemons.  https://review.openstack.org/51155018:42
torgomatictimur: notmyname: ^^18:42
*** vint_bra has quit IRC18:45
*** oshritf has quit IRC18:49
openstackgerritMerged openstack/swift feature/deep: clean up _one_shard_cycle state machine  https://review.openstack.org/51148218:50
*** hseipp has quit IRC18:52
timburkemattoliverau? isn't it like 6am for you? get back in bed!19:06
claygtorgomatic: I'm acctually pretty confused by https://github.com/openstack/swift/blob/master/swift/obj/updater.py#L8919:19
claygIME forking after the eventlet hub has "started" is just *always* bad juju - worse so if the eventlet threadpool think has started!19:19
torgomaticclayg: in this case, we've only got our one greenthread, and we're not using the tpool, so it's not terrible19:20
claygI guess *maybe* somehow the parent process run_forever isn't *really* touching eventlet?  I see a time.sleep at the bottom... so maybe it's just super careful?19:20
mattoliverauYeah it is, and yes I should :p19:20
torgomaticbut yes, fork + eventlet really needs a lot of careful thought19:20
mattoliverauBut now I'm reading about MAMR, thanks notmyname  :p19:21
*** mvk has joined #openstack-swift19:28
openstackgerritTim Burke proposed openstack/swift master: Add Preamble and Postamble to SLO and SegmentedIterable  https://review.openstack.org/36537119:49
*** pcaruana has quit IRC19:58
*** wes_dillingham has quit IRC20:12
timburketorgomatic: thinking a bit about https://review.openstack.org/#/c/510716/ -- what's your opinion on SLOs with *no* segments?20:26
patchbotpatch 510716 - swift - Allow SLOs to have zero-byte last segments.20:26
timburke`curl http://saio:8090/v1/AUTH_test/c/slo?multipart-manifest=put -X PUT -d '[]'` currently works... so i guess we should never break that? do we have tests for it?20:27
openstackgerritMerged openstack/swift feature/deep: Add helper methods to assert shard range equality  https://review.openstack.org/51143720:34
openstackgerritMerged openstack/swift feature/deep: actually use the assertShardRangesNotEqual helper method  https://review.openstack.org/51154420:34
torgomatictimburke: sounds stupid but not worth specifying IMO20:35
*** oshritf has joined #openstack-swift20:35
torgomaticI can easily see how zero-byte final segments would come about, but an empty SLO seems harder to do. At worst, a SLO with a single, zero-byte segment would be created by a naive uploader20:36
*** oshritf has quit IRC20:37
*** vint_bra has joined #openstack-swift20:42
timburkebut i'm right to think that it's a bit late to try *enforcing* that there's at least one segment, yeah?20:43
torgomaticyeah, that ship has sailed20:46
*** vint_bra has quit IRC20:57
*** chetna has joined #openstack-swift21:08
chetnaHi, I wanted to check on GET object API, when we give in valid prefix and valid marker in request we get 204 (curl -i -H "X-Auth-Token: $TOKEN" $STORAGE_URL/container?"prefix=invalid&marker=valid) should we not throw Bad request error code?21:11
torgomaticchetna: what constitutes an invalid prefix? can you give an example of one?21:13
chetna@torgomatic : Anything that does not match any object prefix21:17
torgomaticchetna: that's not really "invalid", it just doesn't match any object right now. One could upload an object starting with that prefix at any time.21:18
torgomaticYou're asking for a set of objects starting with some prefix P; at this time, that set is empty, so the result you receive from Swift is empty. Should that change in the future, you'd receive a non-empty result.21:19
chetnamakes sense.21:19
chetnaThanks for clarifying21:19
torgomaticYou're welcome21:19
*** joeljwright has quit IRC21:21
*** mat128 has quit IRC21:25
torgomaticnotmyname: we're gonna want a backport of patch 511550, assuming it lands21:32
patchbothttps://review.openstack.org/#/c/511550/ - swift - Use "poll" or "selects" Eventlet hub for all Swift...21:32
notmynameyeah21:32
*** wes_dillingham has joined #openstack-swift21:34
*** lifeless has quit IRC21:37
timburketorgomatic: "the old behavior is just so bizarre that I can't imagine anyone wanting it" -- that's *exactly* why i structured that change that way :)21:48
*** catintheroof has quit IRC21:53
*** catintheroof has joined #openstack-swift21:53
*** catintheroof has quit IRC21:55
*** lifeless has joined #openstack-swift22:00
openstackgerritOpenStack Proposal Bot proposed openstack/swift master: Updated from global requirements  https://review.openstack.org/8873622:05
openstackgerritSamuel Merritt proposed openstack/swift master: Shorten typical proxy pipeline.  https://review.openstack.org/50447222:26
*** mvk has quit IRC22:30
openstackgerritSamuel Merritt proposed openstack/swift master: Move all monkey patching to one function  https://review.openstack.org/30661222:42
*** mvk has joined #openstack-swift22:47
*** chetna has quit IRC22:57
*** chetna has joined #openstack-swift22:57
mattoliveraumorning, again..22:58
timburkebetter :-)23:00
*** ChanServ changes topic to "https://bugs.not.mn/project/Swift | Logs: http://eavesdrop.openstack.org/irclogs/%23openstack-swift/ | Meetings: https://wiki.openstack.org/wiki/Meetings/Swift | Review Dashboard: http://not.mn/reviews.html"23:05
-openstackstatus- NOTICE: Workarounds are in place for libcurl and similar dependency errors due to stale ubuntu mirroring, and for POST_FAILURE results stemming from runaway inode utilization on the logs site; feel free to recheck failing changes for either of these problems now23:05
openstackgerritSamuel Merritt proposed openstack/swift master: Remove some leftover threadpool cruft.  https://review.openstack.org/51164423:13
*** mjturek has quit IRC23:15
*** wes_dillingham has quit IRC23:29
*** kei_yama has joined #openstack-swift23:31
*** chetna has quit IRC23:33
*** mvk has quit IRC23:36
*** wes_dillingham has joined #openstack-swift23:37
openstackgerritMerged openstack/swift feature/deep: make root_account, root_container properties of broker  https://review.openstack.org/51150823:42
openstackgerritMerged openstack/swift master: domain_remap: stop mangling client-provided paths  https://review.openstack.org/46421723:42
openstackgerritMerged openstack/swift master: add mangle_client_paths to example config  https://review.openstack.org/51099023:42
openstackgerritMerged openstack/swift master: Imported Translations from Zanata  https://review.openstack.org/51000023:42
*** vint_bra has joined #openstack-swift23:47
*** vint_bra has quit IRC23:52
*** vint_bra has joined #openstack-swift23:53

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