Friday, 2016-09-09

rdogerritrdo-trunk created openstack/swift-distgit: openstack-swift: failed to build 51925e5
*** rpioso has quit IRC00:07
*** aortega has quit IRC00:07
*** nyechiel_ has quit IRC00:11
rdobot[sensu] NEW: - check-delorean-newton-current @ |#| Build failure on centos7-master/current: swift:
*** openstack_tester has quit IRC00:12
rdogerrithguemar proposed openstack/swift-distgit: Add swift.5 manpage
rdobot[sensu] NEW: - check-delorean-newton-head-current @ |#| Build failure on centos7-master-head/current: ironic-inspector, python-tripleoclient:
rdogerritMerged openstack/keystoneauth1-distgit: Use for Source0 URL
rdogerritMerged openstack/keystoneauth1-distgit: Sync requirements
EmilienMthis is why I need puppet packaging ASAP00:34
EmilienMpuppetlabs keeps breaking us00:34
EmilienMnumber80: this one is from today.00:34
rdogerrithguemar proposed openstack/swift-distgit: Add swift.5 manpage
*** saneax is now known as saneax-_-|AFK00:34
number80EmilienM: that's not incorrect, centos-release has the following provides00:43
rdogerritMerged openstack/keystoneclient-distgit: Sync requirements
number80redhat-release = 7.200:44
*** dtrainor has joined #rdo00:44
EmilienMnumber80: have you seen the logs?00:44
EmilienMredhat-release >= 22.0 is needed by puppetlabs-release-22.0-1.noarch00:44
number80EmilienM: they're requiring the wrong min version00:44
EmilienMthat's the bug00:44
number80the bug headline is incorrect as you may think that redhat-release should not be required00:45
EmilienMnumber80: fixed00:46
number80it was likely not even working on RHEL00:46
*** dtrainor has quit IRC00:52
rdogerritrdo-trunk created openstack/oslo-db-distgit: python-oslo-db: failed to build 4e5de54
*** dtrainor has joined #rdo00:53
rdogerrithguemar created openstack/oslo-db-distgit: Create rpm-mitaka branch
rdobot[sensu] NEW: - check-delorean-mitaka-current @ |#| Build failure on centos7-mitaka/current: oslo.db:
EmilienMnumber80: is it related to oslo.db unpin?00:58
*** kenhui has joined #rdo00:59
number80EmilienM: nope, it's a mitaka FTBFS as I needed to fork rpm-mitaka01:01
number80I thought that DLRN could use -rdo branch when available but no01:02
EmilienMoh ok01:03
*** READ10 has joined #rdo01:14
*** pppingme has quit IRC01:28
*** egafford has quit IRC01:31
*** READ10 has quit IRC01:32
*** pppingme has joined #rdo01:34
rdogerritMerged openstack/oslo-db-distgit: Create rpm-mitaka branch
*** thrash is now known as thrash|g0ne01:57
*** akshai has joined #rdo02:04
rdobot[sensu] NEW: - check-delorean-newton-current @ |#| Build failure on centos7-master/current: swift:
*** jhershbe has joined #rdo02:15
*** jhershbe_ has joined #rdo02:20
*** chlong has joined #rdo02:26
*** jhershbe has quit IRC02:28
*** eaguilar has quit IRC02:28
*** jhershbe_ has quit IRC02:28
rdobot[sensu] NEW: - check-delorean-newton-head-current @ |#| Build failure on centos7-master-head/current: ironic-inspector, python-tripleoclient:
*** crossbuilder_ has joined #rdo02:40
*** crossbuilder has quit IRC02:40
*** cwolferh has quit IRC02:49
rdobot[sensu] NEW: - check-delorean-mitaka-current @ |#| Build failure on centos7-mitaka/current: oslo.db:
*** Amii has joined #rdo02:57
*** Amii is now known as Amita02:58
*** fragatina has quit IRC03:03
*** myoung is now known as myoung|afk03:05
*** paragan has joined #rdo03:05
*** pilasguru has joined #rdo03:11
*** sasha2 has quit IRC03:14
*** danielbruno has quit IRC03:28
*** pilasguru has quit IRC03:31
*** stanchan has joined #rdo03:35
*** danielbruno has joined #rdo03:40
*** danielbruno has joined #rdo03:40
*** imcsk8_ has joined #rdo03:42
*** imcsk8 has quit IRC03:42
*** sasha2 has joined #rdo03:48
*** sasha2 has quit IRC03:54
*** rhallisey has joined #rdo03:54
*** kaminohana has quit IRC04:06
rdobot[sensu] NEW: - check-delorean-newton-current @ |#| Build failure on centos7-master/current: swift:
*** sasha2 has joined #rdo04:10
*** absubram has joined #rdo04:16
*** absubram_ has joined #rdo04:16
*** rhallisey has quit IRC04:18
*** paragan has quit IRC04:20
*** paragan has joined #rdo04:20
*** absubram has quit IRC04:20
*** absubram_ is now known as absubram04:20
*** buvanesh_kumar has joined #rdo04:24
*** fragatina has joined #rdo04:26
rdobot[sensu] NEW: - check-delorean-newton-head-current @ |#| Build failure on centos7-master-head/current: ironic-inspector, python-tripleoclient:
*** fragatina has quit IRC04:30
*** akshai has quit IRC04:32
*** coolsvap has joined #rdo04:43
*** iknownothing has joined #rdo04:48
*** snarwade has joined #rdo04:49
*** pgadiya has joined #rdo04:52
*** Son_Goku has quit IRC04:53
rdobot[sensu] NEW: - check-delorean-mitaka-current @ |#| Build failure on centos7-mitaka/current: oslo.db:
*** sasha2 has quit IRC04:56
*** iknownothing is now known as pbronze05:01
*** furlongm_ is now known as furlongm05:01
*** manous has joined #rdo05:01
*** pbronze is now known as iknownothing05:01
*** hynekm has joined #rdo05:07
*** pbandark has joined #rdo05:08
*** Amita has quit IRC05:09
*** jhershbe has joined #rdo05:11
*** jhershbe_ has joined #rdo05:11
*** paragan has quit IRC05:13
*** jaosorior has joined #rdo05:21
*** ganesh has joined #rdo05:22
*** ganesh is now known as gkadam05:22
*** paragan has joined #rdo05:24
*** paragan has joined #rdo05:24
*** chlong_ has joined #rdo05:31
*** jhershbe has quit IRC05:38
*** jhershbe_ has quit IRC05:39
*** richm has quit IRC05:40
*** trishnag has left #rdo05:51
*** aufi has joined #rdo05:59
*** saneax-_-|AFK is now known as saneax06:00
*** aufi has quit IRC06:05
*** aufi has joined #rdo06:05
*** Amita has joined #rdo06:07
rdobot[sensu] NEW: - check-delorean-newton-current @ |#| Build failure on centos7-master/current: swift:
*** ccamacho has joined #rdo06:13
rdogerritrdo-trunk created openstack/swift-distgit: openstack-swift: failed to build 51925e5
*** pcaruana has joined #rdo06:23
*** iranzo has joined #rdo06:27
rdobot[sensu] NEW: - check-delorean-newton-head-current @ |#| Build failure on centos7-master-head/current: swift, ironic-inspector, python-tripleoclient:
*** jprovazn has joined #rdo06:29
*** chandankumar has joined #rdo06:30
*** manous has quit IRC06:32
*** sasha2 has joined #rdo06:33
*** fzdarsky has joined #rdo06:35
*** smeyer has joined #rdo06:36
*** chandankumar has quit IRC06:39
*** sasha2 has quit IRC06:42
*** vaneldik has joined #rdo06:42
rdogerritMerged openstack/tempest-distgit: Added python-tempest, python-tempest-tests and openstack-tempest-all subpackage
*** vaneldik has quit IRC06:43
*** manous has joined #rdo06:45
*** satya4ever has joined #rdo06:47
*** chandankumar has joined #rdo06:52
*** vaneldik has joined #rdo06:52
*** mosulica has joined #rdo06:54
rdobot[sensu] NEW: - check-delorean-mitaka-current @ |#| Build failure on centos7-mitaka/current: oslo.db:
*** zoli_gone-proxy is now known as zoliXXL06:58
*** zoliXXL is now known as zoli|wfh06:58
*** sasha2 has joined #rdo06:59
*** florianf has joined #rdo07:00
*** social has joined #rdo07:06
*** tesseract- has joined #rdo07:07
*** flepied has quit IRC07:07
*** zoli|wfh is now known as zoli_gone-proxy07:11
*** mcornea has joined #rdo07:12
*** zoli_gone-proxy is now known as zoliXXL07:17
*** jpena|off is now known as jpena07:18
*** chlong has quit IRC07:19
*** chlong_ has quit IRC07:19
*** ihrachys has joined #rdo07:20
jpenagood morning07:20
*** smeyer has quit IRC07:27
*** smeyer has joined #rdo07:27
*** rgogunskiy has joined #rdo07:29
*** jpich has joined #rdo07:29
kashyaprbowen: garrett: Thanks, both.  Just caught up with scroll, saw the 'mw2md' repo.07:30
*** pnavarro has joined #rdo07:33
*** limao has joined #rdo07:34
*** limao_ has joined #rdo07:37
*** KarlchenK has joined #rdo07:39
*** limao has quit IRC07:39
*** k4n0 has joined #rdo07:40
*** flepied has joined #rdo07:42
*** nyechiel_ has joined #rdo07:43
*** ohamada has joined #rdo07:47
rdogerritMerged openstack/swift-distgit: Add swift.5 manpage
*** mbound has joined #rdo07:54
*** gildub has quit IRC07:56
*** jlibosva has joined #rdo07:58
*** mbound has quit IRC07:59
*** derekh has joined #rdo08:05
*** dsneddon has quit IRC08:06
*** athomas has joined #rdo08:06
rdobot[sensu] NEW: - check-delorean-newton-current @ |#| Build failure on centos7-master/current: swift:
*** lmiccini_ has joined #rdo08:09
*** firat has joined #rdo08:10
*** mflobo has joined #rdo08:10
*** lmiccini has quit IRC08:10
*** snecklifter has joined #rdo08:19
*** lmiccini_ is now known as lmiccini08:23
*** firat has quit IRC08:23
*** dgurtner has joined #rdo08:25
*** dgurtner has joined #rdo08:25
*** nyechiel_ has quit IRC08:25
rdobot[sensu] RESOLVED: - check-delorean-newton-current @ |#| No build failures detected:
*** karimb has joined #rdo08:28
rdobot[sensu] NEW: - check-delorean-newton-head-current @ |#| Build failure on centos7-master-head/current: swift, ironic-inspector, python-tripleoclient:
*** KarlchenK has quit IRC08:36
*** jaosorior is now known as jaosorior_lunch08:40
rdogerritJakub Ruzicka proposed rdopkg: patch: fix traceback with no new patches
*** jlibosva has quit IRC08:43
rdogerritMerged rdopkg: patch: fix traceback with no new patches
*** jschlueter has quit IRC08:45
*** jjoyce has quit IRC08:45
*** myoung|afk has quit IRC08:45
derekh2016-09-09 06:10:24.033740 | 2016-09-09 06:10:23.000 | [Errno 14] HTTP Error 302 - Found08:51
derekhhmm, thats not what I wanted to paste08:52
*** _degorenko|afk is now known as degorenko08:52
derekhthis and others like it, is now being redirected too ->
derekhhas this just changes?08:53
*** imcsk8_ has quit IRC08:54
rdobot[sensu] NEW: - check-delorean-mitaka-current @ |#| Build failure on centos7-mitaka/current: oslo.db:
*** tosky has joined #rdo08:55
*** gszasz has joined #rdo08:56
*** imcsk8 has joined #rdo08:56
*** akrivoka has joined #rdo08:57
*** cwolferh has joined #rdo08:58
rdogerritMerged config: Add missing projects to the zaqar integrated template
rdogerritJavier Peña proposed rdoinfo: Set Tempest branches for newton-uc and mitaka
*** sdake has joined #rdo09:02
*** lkuchlan_wfh has joined #rdo09:02
*** kbyrne has quit IRC09:04
*** abehl has joined #rdo09:04
*** Son_Goku has joined #rdo09:05
*** nijaba has quit IRC09:05
*** nijaba has joined #rdo09:06
*** nijaba has quit IRC09:06
*** nijaba has joined #rdo09:06
chandankumarjpena: dmellado number80 \o/09:07
chandankumarjruzicka: \o/09:08
derekhdelorean-deps here
derekhis adding this repo
derekhthen all rpm downloads in that directory get a 302 to here
derekhcould we just get the cdn version in delorean-deps.repo?09:09
*** zoliXXL is now known as zoli|wfh09:09
*** kbyrne has joined #rdo09:10
*** stanchan has quit IRC09:11
rdogerritsahid proposed openstack/nova-distgit: add package for placement API service
*** shardy has joined #rdo09:11
derekhWe've been redirected during yum installs 10,000 times in the last 12 hours09:12
*** limao_ has quit IRC09:12
*** limao has joined #rdo09:12
dmelladohey chandankumar o/09:12
derekhsshnaidm|afk: ^^09:13
jpenakbsingh, is the above redirect ^^ a permanent one? We can update our deps file to avoid double connections09:14
*** mbound has joined #rdo09:14
kbsinghjpena: whst is the problem?09:16
*** sdake_ has joined #rdo09:16
jpenakbsingh, trying to fetch a package from results in a 302 redirect to (note the additional 'cdn')09:16
kbsinghis that a problem?09:17
jpenawhen installing tons of packages, it's better to go straight to buildlogs.cdn then. Just wondering if buildlogs.cdn is a safe url to use09:17
*** jlibosva has joined #rdo09:17
kbsinghyou will see another 302 after that09:17
kbsinghno. not safe09:17
jpenaaha, ok. derekh ^^09:17
*** paragan has quit IRC09:18
kbsinghthere are about 150 machines on the cdn edge09:18
*** sdake has quit IRC09:18
kbsinghand the service level is that at any point in time ~ 30 are active09:19
derekhkbsingh: but doesn't the dns entry stay the same?09:19
derekhkbsingh: I'm trying to avoid 1,000's of these per hours09:20
derekh1472978811.072    364 TCP_MISS/302 718 GET - HIER_DIRECT/ text/html09:20
derekh1472978811.277    102 TCP_MEM_HIT/200 16697 GET - HIER_NONE/- application/x-rpm09:20
kbsinghit uses the first redirect to workout where you are09:20
kbsinghderekh: for the edge yes. and the dns server uses the A lookup inbound09:20
jruzickajpena, I'm testing info-tags-diff and I wonder if it's wise that it sometimes returns [] and sometimes {}09:20
*** limao has quit IRC09:20
jruzicka+     openstack-banana ['newton-uc', 'newton']09:20
jruzicka+     openstack-panko {'under-review': None}09:20
kbsinghderekh: why are you trying to save them? over 10k rpm' the redirect is likely not a huge time and data overhead09:21
number80jruzicka: no09:21
number80but shouldn't we fix the format instead/09:21
jruzickanumber80, how? :)09:22
jruzickawhich format?09:22
jruzickajpena, you only need list of changed tags, no?09:22
jpenajruzicka, we might want to fix formatting, but there's a reason. For openstack-banana, there was a change in 2 branches, for openstack-panko, it's a brand new package09:22
jruzickajpena, ah, OK so it's intended09:22
jruzickaneed to mention that in the doc, then09:22
number80jruzicka: using an under-review key09:23
number80the rdo.yml one09:23
jruzickajpena, just to be sure... only changes/instertions are shown, deletion isn't indicated in any way09:24
jruzickathat's also intended09:24
jpenammm... haven't thought of that09:24
derekhkbsingh: because I've been seeing a bunch or yum installs fail this morning with messages about 302's09:24
kbsinghderekh: how did they fail?09:25
number80 /buffer 709:25
jpenajruzicka, for our use case we don't care about removals, it's just packages that no longer need to be built.09:25
*** apevec has joined #rdo09:26
derekhkbsingh: e.g.
derekhkbsingh: 2016-09-09 08:02:25.755724 | 2016-09-09 08:02:06.000 | [Errno 14] HTTP Error 302 - Found09:26
jruzickajpena, yes, that's OK. I'm just making sure I'm not documenting bad behavior ;)09:26
kbsinghand this worked yesterday?09:26
apevecderekh, this redirect to cdn09:27
apevecwhy is that error?09:27
derekhkbsingh: its the first time I've noticed it, not all jobs are failing09:27
apevecderekh, curl is not following redirect by default09:28
apevecbut yum should09:28
apevec(curl CLI that was)09:28
kbsinghlet me.confirm there is no change on our side in the last 24hs09:28
apevecderekh, in which part of code is that error reported?09:29
derekhapevec: kbsingh it could possibly be a network problem our side, and maybe the 302 is reported because that was what yum origionaly saw09:29
apevecah that's from yum09:29
derekhapevec: ya, its a yum install of python-tripleoclient
jruzickajpena, finally, when a package config is changed, but no tags changes are present, openstack-tempest [] is returned. Is that also OK or did I accidentaly change behavior of that?09:31
apevec#define EFAULT      14  /* Bad address */09:31
apevecderekh, might be DNS temp failure?09:31
apevecrecheck it09:31
kbsingh97% of all buillogs cdn traffic is rdo09:32
jpenajruzicka: that happens when the list of maintainers is changed... We should take it into account, in that case we should not return anything09:32
apeveckbsingh, we'll get images09:33
apevecout of there09:33
kbsinghderekh: no change in this side09:33
derekhkbsingh: apevec maybe I'll chase down possible network problems on our end (we have been having some), but I still don't understand what we can have baseurl=http://buildlogs.cdn .... to avoid the redirect, the multiple machines behind this DNS entry arn't visible to the user09:33
derekhkbsingh: thanks for checking09:33
kbsinghapevec: this is not a problem. we stood up the cdn specifically for rdo09:33
*** Alex_Stef has joined #rdo09:33
apeveckbsingh, can you get us traffic break-down per folder ?09:34
kbsinghderekh: the dns server uses the inbound A lookup for geo location09:34
apevecI wonder how much of that is images vs RPMs09:34
kbsinghapevec: prolly09:34
kbsinghapevec: i know we hit 3 to 3.5gbit at sometimes during the day09:34
apevecderekh, redirects is not causing this09:35
kbsinghderekh: btw is all this coming from the sameIP.. ie out of a NAT gateway?09:35
apevecalso buildlogs.cdn is not complete mirror09:36
apeveconly .rpm is mirrored afaik09:36
*** lkuchlan_wfh has quit IRC09:36
kbsinghthere is no repodata or md5 etc on cdn09:36
apeveckbsingh, thanks for confirming09:36
kbsinghso a compromise on a node will still have some level of safety09:36
derekhkbsingh: yup if I request a file at directly you get a inbound A request, the fact that I redirected from is irrelevant is it not ?09:36
derekhkbsingh: it would all be the same IP, its going through a proxy09:37
kbsinghderekh: the redirect is used too09:37
*** nyechiel_ has joined #rdo09:38
kbsinghhow would you do this.. change the urlgrabber to support aliases?09:38
derekhahh, we need the redirect anyways because of the absence of repodata09:40
derekhits got repodata
rdogerritJakub Ruzicka proposed rdopkg: info-tags-diff: new utility action to show rdoinfo tag changes
jruzickajpena, once that is merged I'll release new version.09:44
jruzickanot sure about releasing on friday but hey, at least the breakage will have enough time to propagate :-p09:45
number80apevec: found a new issue w/ automation => git merge ate my subpackage09:46
*** nyechiel_ has quit IRC09:46
number80in some case, adding a new subpackage and git merge may think it is an old one that was modified09:46
kbsinghderekh: right, so we control the metadata, that way a compromise in the CDN and content changing there should fail, since it wont match metadata09:46
* number80 pessimistic that we can ever fully automate that part09:47
jruzickanumber80, you mean optimistic about machines not taking over? :D09:47
derekhkbsingh: ok, that makes sense09:47
kbsinghderekh: there are 2 challenges we are working around here (1) the content on buildlogs isnt signed. so we need the shasum's from the metadata to confirm authenticity ( well, confirm is a strong workd - but build confidence might be more accurate ) and (2) we want rapid promotion from cbs.centos.org09:47
number80jruzicka: I wish they do if we can turn me into a flame-throwing giant robot :)09:48
*** oshvartz has quit IRC09:48
kbsinghalso, was on the phone earlier- back to the laptop, so can actually type again - w.r.t the redirect, its used by the geo distribution side, so it needs the A lookup, to workout what AS you are coming out of and then the redirect to workout what content. eg. I'm restricting the number of machines in the CDN that get the tripleO images to make that promotion faster09:49
jruzickanumber80, yeah, me too, but don't tell anyone09:49
kbsinghso we could have ~ 50 machines that proxy for the rpms, but only 5 that do it for the images. and usually this starts wit 8 machines, and it expands based on demand per object ( rpm or image )09:49
kbsinghhaving said all that <insert hand waving> - if all your content is entire in and out of a single IP, and wont be any exceptions, you can take some liberties. But overall best to just leave it alone and let the complexity be handled on the remote end09:50
apevecnumber80, do you have an example review?09:51
apevecwhere merge was a hungry little boy09:51
jruzickaapevec, btw
number80apevec: it happened during fedora sync (not affecting RDO)09:53
*** nyechiel_ has joined #rdo09:53
*** Son_Goku has quit IRC09:53
apevecjruzicka, thanks for writing that, I've linked in the review that you have linked there :)09:54
*** akrivoka has quit IRC09:54
apevecomg, hyperlinking loop!09:54
jruzickaoh my god, I hope internet survives that09:54
jruzickaapevec, also, once we get lots of patches, incoming queue will get messy on review.rdo*09:55
jruzickabecase they're never closed from gerrit POV09:55
*** akrivoka has joined #rdo09:56
jruzickawe might need to hack rpmfactory in future09:56
derekhkbsingh: ok, I understand the security implications and wanting to control the metadata, and it makes sense and is a good enough reason alone to have the redirects09:56
* jruzicka is awake before 12:00 for several days straight, what a strange world09:57
*** sdake has joined #rdo09:57
derekhkbsingh: I just don't get how it effects what hosts are serving my requests, but thats fine I'll do a bit a reading up on it later09:57
apevecjruzicka, !09:58
* apevec cannot have haircut everyday 8-9am09:58
apevecsimply not enough material :)09:58
* jruzicka lol'd09:59
derekhkbsingh: one more thing Re.
apevecderekh, redirect should all just work, provided networking/DNS etc. is working09:59
*** sdake_ has quit IRC10:00
derekhkbsingh: just want to confirm you were asking about image builds that are running on centos infrastructure? not images builds that download from centos infrastructure10:00
apevecderekh, kbsingh - for that thread we need weshay_pto to chime in (once s//_pto/ )10:00
apevecderekh, it's about RDO CI10:01
apevecah he did chime in10:01
derekhapevec: yup, I know it should all work, the conversation had continued because I was trying to understand how it was useful10:01
derekhapevec: ok, thanks10:02
*** nyechiel has joined #rdo10:04
kbsinghderekh: so the tl;dr is - kinda sorta.10:06
kbsinghderekh: the longer version is that a lot of ( all of ? ) the infra for the management of, the admin of, and the control of is virtualised, and were trying to make sure its always in the best state10:06
jpenaapevec: we have a funny bug in the rdo-release package for the test day. We're shipping /etc/pki/rpm-gpg/RPM-GPG-KEY-CentOS-SIG-Virtualization-RDO, but the qemu-ev repo file included in the package has RPM-GPG-KEY-CentOS-SIG-Virtualization as the key name10:06
kbsinghderekh: the specific challenge we had was that every 1byte of image movement inside the ci infra, was causing 6 bytes of data to move around10:07
*** nyechiel_ has quit IRC10:07
*** KarlchenK has joined #rdo10:07
apevecjpena, send PR please10:08
jpenaapevec: on it, first I'm adding a note to the workarounds page10:08
*** richm has joined #rdo10:08
apevecI missed that in number80's PR10:08
apevecit was renamed to avoid conflict with centos-release-*rpm10:08
apevecI'll release update asap10:09
*** jprovazn has quit IRC10:11
*** Son_Goku has joined #rdo10:12
derekhkbsingh: ya, I'm sure your dealing with a LOT of data moving around10:19
*** KarlchenK has quit IRC10:20
*** shardy has quit IRC10:21
*** Son_Goku has quit IRC10:23
*** k4n0 has quit IRC10:26
*** sshnaidm|afk is now known as sshnaidm|pto10:28
rdobot[sensu] NEW: - check-delorean-newton-head-current @ |#| Build failure on centos7-master-head/current: ironic-inspector, python-tripleoclient:
apevecnumber80, we might want to update python-PyMySQL, 0.6.7 in common vs 0.7.6 in upstream u-c10:29
apevecI noticed b/c tonyb was testing 0.7.7
*** thrash|g0ne is now known as thrash10:30
*** sdake_ has joined #rdo10:31
*** sdake has quit IRC10:33
chandankumarjpena: Does new openstack-tempest rpms are available in rdo-release-newton.rpm?10:35
jpenachandankumar: no, they were built after the last CI pass10:36
*** Guest90983 is now known as rook10:36
chandankumarjpena: any plan to pull there/10:37
jpenaas soon as we get another CI success :)10:37
*** nyechiel has quit IRC10:38
*** Son_Goku has joined #rdo10:38
*** rhallisey has joined #rdo10:39
*** Son_Goku has quit IRC10:40
*** jaosorior_lunch is now known as jaosorior10:41
*** leanderthal|afk is now known as leanderthal10:41
apevecjpena, chandankumar I've disabled promotion pipeline during testday10:41
*** paragan has joined #rdo10:41
apevecwe'll turn it on later today10:41
apevecchandankumar, in the meantime it's in
apevecnumber80, jpena - passed CI b/c CI job is using master not master-head right?10:46
apevecwhat shall we do we thos master-head FTBFS ?10:46
jpenaapevec: right10:46
apeveccan we move them to the separate topic?10:46
apevecjpena, is topic configurable?10:47
apevecper DLRN instance10:47
apevecI'd like to keep track of them but lower prio10:47
jpenaapevec: right now we have . We could add some logic10:47
apevecflepied, ^ what say ya?10:47
apevecI'll rename topic on existing master-head ftbfs10:48
apevecrdo-FTBFS-head ?10:48
*** gildub has joined #rdo10:50
chandankumarIs anyone hitting this error while using w10:53
apevecchandankumar, which error?10:53
chandankumar [Errno 14] HTTP Error 404 - Not Found10:54
rdobot[sensu] NEW: - check-delorean-mitaka-current @ |#| Build failure on centos7-mitaka/current: oslo.db:
apeveckbsingh, ^ have we lost content on mirror.c.o ?10:55
apeveckbsingh, ^10:55
chandankumaropenstack-newton not found.10:55
apevecchandankumar, newton is not published yet :)10:55
apevecyou have enalbed release repo10:55
*** athomas has quit IRC10:56
apeveckeep only testing enabled as it is by default10:56
apevecchandankumar, but on a clean box you'll hit issue fixed by jpena
apevecI'm uploading this update now10:56
apevecjpena, where have you put a note about rdo-release bug?10:59
apevecah it's not published yet, you can cancel it now :)11:01
apevechmm, looks like we have a backlog there11:01
*** buvanesh_kumar has quit IRC11:02
*** athomas has joined #rdo11:02
apevecchandankumar, ah that was your issue, you need CONFIG_ENABLE_RDO_TESTING=y11:03
*** jkilpatr has quit IRC11:04
*** pkovar has joined #rdo11:04
*** matbu is now known as matbu|lunch11:05
rdogerritMerged openstack/gnocchi-distgit: jsonpatch is a new dep of Gnocchi
*** egafford has joined #rdo11:07
*** karimb has quit IRC11:08
chandankumarapevec: done, Thanks !11:08
*** pkovar has quit IRC11:09
*** akshai has joined #rdo11:11
*** egafford has quit IRC11:13
*** ohamada has quit IRC11:15
*** ohamada has joined #rdo11:15
*** pkovar has joined #rdo11:15
*** Son_Goku has joined #rdo11:15
*** absubram has quit IRC11:16
*** lkoranda has quit IRC11:16
*** lkoranda has joined #rdo11:17
*** Alex_Stef has quit IRC11:23
*** fultonj has joined #rdo11:27
*** spr has quit IRC11:28
*** jschlueter has joined #rdo11:28
*** alexismonville has joined #rdo11:29
*** jhershbe has joined #rdo11:31
*** jhershbe_ has joined #rdo11:31
*** mburned_out is now known as mburned11:31
*** zoli|wfh is now known as zoli|lunch11:32
*** alexismonville has quit IRC11:34
flepiedapevec: fine with me. we should also fix the passing of the GERRIT env variable that is not done.11:36
*** Son_Goku has quit IRC11:36
*** jprovazn has joined #rdo11:36
*** jkilpatr has joined #rdo11:38
*** nyechiel has joined #rdo11:38
*** dpeacock has joined #rdo11:39
*** karimb has joined #rdo11:42
*** alexismonville has joined #rdo11:46
*** abregman has joined #rdo11:47
*** Son_Goku has joined #rdo11:47
*** jprovazn has quit IRC11:50
*** pkovar has quit IRC11:51
*** kenhui has joined #rdo11:53
*** lucasagomes is now known as lucas-hungry11:55
*** nyechiel has quit IRC11:55
flepiedI tried to reproduce the minimal_pacemaker issue we have in the CI but it seems to work fine on my test machine...11:55
*** Amita has quit IRC11:55
*** pkovar has joined #rdo11:56
rdogerritJavier Peña created DLRN: Use different review topics depending on the baseurl
*** jprovazn has joined #rdo11:57
*** Son_Goku has quit IRC11:58
apevecflepied, yep, same was for trown - worked on his minidell11:59
apevecflepied, so it's really stress test by slow HW in ci.centos11:59
flepiedapevec: yes it seems11:59
apevecand still important to look at closely by gnocchi/tripleo11:59
*** gkadam has quit IRC12:00
flepiedapevec: do we have some tools to display graphs from the dstat data?12:01
*** ohochman has quit IRC12:01
*** trown|outtypewww is now known as trown12:01
flepiedtrown: ^12:01
apevecwhich dstat data?12:02
trownI dont think we have dstat data on the overcloud12:02
flepiedtrown: you are right it's only on the undercloud12:02
trownovercloud would need to be setup as part of THT12:03
trownwould be a good RFE for tripleo though12:03
flepiedit could be handy to know what is the minimal requirements for machines12:03
apevecand where is dstat on the undercloud?12:03
trownhaving some better performance data for the deploy would be nice12:03
trownhmm... thought we had it on the undercloud... dont see that either12:05
snarwadeapevec, can you please review ?12:05
EmilienMtrown: it's more something for monitoring with opstools12:05
EmilienMdtstat is not a monitoring tool, it's more analyze12:06
EmilienMto watch metrics like cpu, memory, etc, sensu will do it12:06
trownya being able to correlate which processes are consuming those would be really useful12:07
trownright now it is pretty much guessing based on /var/log/messages/12:07
*** karimb has quit IRC12:07
*** jayg|g0n3 is now known as jayg12:08
*** jpena is now known as jpena|lunch12:12
chandankumarapevec: i need to include python-gabbi as a dependency in tempest12:12
chandankumarotherwise testr list-tests will fail12:13
chandankumarand we need to install python-aodh-tests also12:13
EmilienMwe did it for python gnocchi tests12:14
*** matbu|lunch is now known as matbu12:15
*** dneary has joined #rdo12:15
apevecchandankumar, just send review :)12:23
chandankumarEmilienM: ok12:23
* chandankumar will be back in an hour.12:26
*** chandankumar is now known as chkumar|out12:26
apeveccoolsvap, your're original author of Using GlusterFS for Cinder with RDO ?12:28
*** pradk has joined #rdo12:28
apevecplease review change from snarwade ^12:28
EmilienMapevec: puppet promoted last night and just right now fyi12:28
*** jhershbe has quit IRC12:28
apevecEmilienM, cool, much promotions!12:28
apevecEmilienM, so that's w/ unpinned keystone?12:28
rdobot[sensu] NEW: - check-delorean-newton-head-current @ |#| Build failure on centos7-master-head/current: ironic-inspector, python-tripleoclient:
coolsvapapevec: will do12:29
*** jhershbe_ has quit IRC12:29
EmilienMapevec: yes12:29
apevecEmilienM, excellent, did tripleo-ci periodic also run over the night with latest repo?12:30
EmilienMapevec: I haven't checked yet, let me see12:30
EmilienMit doesn't sounds like12:30
apevechmm sshnaidm|pto weshay_pto ^ ?12:30
EmilienMsshnaidm|pto is not here, he usually checks12:30
EmilienMeveryone on pto!12:30
apevecderekh, trown, ^ do you know is tripleoci periodic is disabled ?12:31
EmilienMperiodic-tripleo-ci-centos-7-ovb-ha success12:31
EmilienMperiodic-tripleo-ci-centos-7-ovb-nonha failure12:31
EmilienM2016-09-09 06:14:20.791335 | 2016-09-09 06:14:20.000 |   jq-1.3-2.el7.x86_64: [Errno 256] No more mirrors to try.12:31
apevecwhich repo hash ?12:31
EmilienMI don't know it use
*** Son_Goku has joined #rdo12:33
apevecEmilienM, where are the logs?12:33
EmilienMapevec: bookmark
*** Amii has joined #rdo12:34
rdogerritjason guiditta created puppet/puppet-barbican-distgit: Add spec for puppet-barbican.
*** iknownothing has quit IRC12:35
*** zoli|lunch is now known as zoli|wfh12:36
EmilienMapevec: heh it's morning :-(12:36
apevecthat's 2016-09-08 23:5812:36
EmilienManyway, it was transient12:36
EmilienMlet's recheck it12:36
EmilienMif I find how to do ti12:37
apevecwith keystone build  20160905112812:37
apevecso still pinned12:37
rbowenGood morning #rdo12:37
apevecoops, typed and it does exist :)12:38
EmilienMrecheck on
EmilienMlet's see if it trigger periodic job12:38
EmilienMok I see the jobs in zuul12:38
EmilienMlet's see what repo it does use now and if it pass :)12:39
EmilienMi'll monitor that12:39
apevecthat's Wes' evil trick :)12:39
dmsimarddamn, rbowen beat me to hello #rdo12:41
rbowenachievement unlocked!12:41
dmsimardmisc: you around ?12:42
EmilienMdmsimard: we promoted, you can try again your magnum game12:42
dmsimardhaha, ok12:42
dmsimardEmilienM: wait, you promoted beyond yesterday's patch ?12:43
dmsimardEmilienM: or do I need to promote and then test magnum12:43
*** gildub has quit IRC12:43
ayoungLooking at  I see src and doc rpms for python-keystoneauth1, but not the binary.  What am I missing?12:43
ayoungah python2-12:44
snarwadethanks apevec coolsvap rbowen12:45
*** aortega has joined #rdo12:46
apevecayoung, yes, complain to fedora packaging comitee :)12:47
*** sdake_ is now known as sdake12:47
ayoungapevec, nah, it is a good approach, just caught me off guard12:47
ayoungmeans we have a path to py312:47
apevecayoung, also bookmar /consistent/12:47
rdogerritAlexey Deryugin created openstack/murano-distgit: Add murano-tempest-plugin extra files to package
ayoungapevec, I actually want the lates to check something12:48
apevecayoung, it's not good for backward compat, rpm Provides sometimes do not work12:48
apevecthey should've kept python- for py212:48
ayoungapevec, I was testing kerberos support, and that is a little wonky.  It uses optional, and the error message referst to pip12:48
rdobot[sensu] RESOLVED: - check-delorean-mitaka-current @ |#| No build failures detected:
rdogerritjason guiditta created puppet/puppet-ovn-distgit: Add spec for puppet-ovn.
*** ohochman has joined #rdo12:48
ayoungapevec, I'll take your word on it.  Its one of those language things that will burn you one way or the other12:49
apevecayoung, atm current == consistent but safer to start from consistent to ensure you don't have packages out of sync due to ftbfs12:49
EmilienMdmsimard: promotion was automatic this night12:49
dmsimardEmilienM: yeah I just saw
EmilienMsemi auto I mean12:49
dmsimarddmellado: you sure merged that tempest change quickly12:51
rdogerritrdo-trunk created openstack/tempest-distgit: openstack-tempest: failed to build 0f8baae
dmsimarddmellado: I bet some QE/CI folks don't see that one coming12:51
*** lucas-hungry is now known as lucasagomes12:51
dmsimardtrown: you're using openstack-tempest in oooq right ?12:51
dmsimardthe rpm ?12:51
dmsimard(and so does tripleoci?)12:52
*** jaosorior has quit IRC12:52
*** unclemarc has joined #rdo12:52
dmsimardI'll create rpm-mitaka branch for tempest12:52
*** jaosorior has joined #rdo12:53
dmsimardhrm, wait, there's already rpm-mitaka12:53
*** rlandy has joined #rdo12:54
*** Goneri has joined #rdo12:54
dmsimardWhy are we seeing a ftbfs in centos7-mitaka for a change that occured only in rpm-master ? It wants to pull in designate, horizon and watcher test plugins which do not exist in mitaka.12:55
dmsimardapevec, jpena|lunch ^12:55
apevecdmsimard, which one?12:55
dmsimardthe projects.ini has the right config12:58
dmsimardrdoinfo ?
*** jcoufal has joined #rdo12:59
dmsimardmitaka has source-branch master12:59
apevecdmellado, ^ need to switch to ?13:00
rdobot[sensu] NEW: - check-delorean-mitaka-current @ |#| Build failure on centos7-mitaka/current: tempest.git:
*** pgadiya has quit IRC13:01
dmsimardapevec: would the fact that source-branch is set to master in rdoinfo make it so centos7-mitaka pulls rpm-master instead of rpm-mitaka ?13:01
dmsimardif so, it's going to be a problem when we move to upstream tempest since it's not branched13:01
dmsimardI believe there is a distro-source-branch macro or something like that ?13:02
rdogerritAlexey Deryugin proposed openstack/murano-distgit: Add murano-tempest-plugin extra files to package
apevecit should not, not sure what happend - but source-branch needs to be adjusted13:02
dmelladodmsimard: apevec13:02
dmelladotempest is branchless but the plugins *are* branched13:02
dmelladothat's why we were artificially branching that13:02
apevecthose in-tree13:02
dmsimarddmellado: right now mitaka is building off of master13:03
dmelladodmsimard: jpena|lunch had a patch for that13:03
dmsimardhe did ?13:03
* dmsimard looks13:03
dmelladoyep, let me check for it13:03
apevecmitaka: This branch is 163 commits ahead, 1081 commits behind openstack:master.13:03
dmsimardjpena|lunch is so good
dmsimardmy hero13:03
dmelladoapevec: that's what eggmaster and me put for osp13:04
*** ohochman has left #rdo13:04
*** rgogunskiy has quit IRC13:04
dmelladoin any case, if you prefer to build off master while dealing with the plugin issues13:04
dmelladothat'd be another option13:04
*** sdake has quit IRC13:05
apevecmaster-head will build from master13:05
kashyapajo: Sometime soon probably you should try to get 'os-log-merger' as a proper Fedora / RDO RPM --
ajoping apevec, how do I get a new project created for rrdoproject?  ? :)13:05
apevecdmellado, how is kept up to date?13:06
ayoungapevec, they did something with keystoneauth1:  they put in extras, and the kerberos federated plugin is in there.  I have the latest python2-keystoneauth1 rpm installed, but epi does not show that group13:06
ajokashyap, we have telepathy :P13:06
dmelladoapevec: that's sync with latest tagged version13:06
kashyapajo: Hehe13:06
apevecayoung, epi?13:06
dmelladogreat tool, check it if you don't know it kashyap13:06
ayoungvery useful tool13:06
dmelladoapevec: there'd be most prob another tag in one week13:07
kashyapdmellado: I saw ajo demo it live, and also forced him to rename it to "os-log-merger" as opposed to "oslogmerger" (Oslo G Merger)13:07
dmelladoor so, after the mid-cycñe13:07
apevecajo, if that's generic enough, put it in Fedora13:07
ayoungapevec, the setup.cfg for keystoneauth looks like this13:07
dmelladokashyap: I meant epi xD13:07
kashyapapevec: That was what I was suggesting, it could be generic13:07
apevecif specific to openstack services (which we removed from Fedora) add it to RDO13:07
kashyapdmellado: Ah, /me looks13:07
*** _elmiko is now known as elmiko13:08
ayoungSo I am not sure what happens with is a python thing13:08
ajoapevec, it's quite openstack specific (openstack log merger)13:08
ajobut it's not an openstack hosted project (yet)13:08
ayoungepi group show  keystoneauth1.plugin  show v3fedkerb               | keystoneauth1.extras.kerberos._loading          | MappedKerberos                 | keystoneauth1 2.12.1         |13:09
ayoungSo it is installed, but running OSC gives:13:09
ajoso, add it to rdo, any pointer? shall I bring just the topic on the list or make the request somewhere?13:09
apevecajo, is it changing a lot? I.e. do you want it in trunk repo chasing your master branch?13:09
apevecajo, it's
dmelladoajo: how about adding it to openstack first?13:09
apevecbut yeah what dmellado said :)13:09
*** jprovazn has quit IRC13:10
apevecyou could just host it (ala stackforge) at review.o.o first13:10
ajoapevec, not changing a lot, slow moving yet, I may want to integrate it with logstash soon13:10
ajobut that's still not there13:10
ayoungNoMatchingPlugin: The plugin v3fedkerb could not be found13:10
ajoapevec, I must admit I started that once, failed and didn't try again, I will do it13:10
*** karimb has joined #rdo13:11
apevecdmsimard, jpena - what do you think about slow-moving projects? They're not worth to bother with RDO trunk13:11
dmsimardapevec: slow moving projects ?13:11
apevecbut could be worth hosting in common part of RDO repo13:11
apevecdmsimard, like, not worth chasing master13:12
apevecand not worth causing us FTBFS :)13:12
dmsimardapevec: example ?13:12
apevecdmsimard, from ajo ^13:12
apevechelper tool13:12
apevecfor log analysis13:12
EmilienMpyparsing-2.0.3-1.el7.noarch: [Errno 256] No more mirrors to try.13:13
ajowhy do you care about if it's slow or fast moving ?13:13
dmsimardwell first it'd have to be in openstack namespace, I don't think we want to chase non-openstack projects at all -- these would go in -deps instead (and thus not chased)13:13
ajoI'm a bit lost, sorry :)13:13
EmilienM is so unreliable13:13
*** kenhui has quit IRC13:13
ajodmsimard, apevec moving to openstack namespace makes sense13:14
*** nyechiel has joined #rdo13:14
ajo"apevec: moving"13:14
derekhEmilienM: I've been seeing similar errors all day13:14
ajonot moving apevec to the openstack namespace,13:14
EmilienMderekh: :( it prevents us to promote13:14
pabelangerEmilienM: is that where DLRN builds end up?13:15
EmilienMpabelanger: no13:15
dmsimardpabelanger: no.13:15
apevecEmilienM, that's smoething derekh was looking at earlier, what is the exact error?13:16
EmilienMapevec: what I pasted13:16
EmilienMit fails to download pyparsing13:16
EmilienMand boom13:16
apevecpabelanger, it's CBS builds13:16
dmsimardapevec: and then, second, I'm not sure we want to package "tools" that aren't dependencies of openstack projects we're packaging ? Would we package ARA because it's in /openstack ? Shade I can understand, but it still falls in the same category ?13:16
pabelangerapevec: dmsimard: at what point do they end up on
apevecright, so not Trunk-worthy as I said13:17
dmsimardin the end, we have to ask ourselves -- do we package what /we/ need, or what the /users/ need ?13:17
apevecpabelanger, once tagged -release13:17
apevecbuildlogs is -testing13:17
apevecinitial CBS koji build is -candidate13:17
pabelangermaybe we could mirror buildlogs to openstack-infra AFS13:18
dmsimardnumber80: hey, you created the puppet-barbican stuff ?13:18
apevecdmsimard, so where do put such tools, should we create new repo or just put it in common ?13:18
dmsimardgerrit project/distgit/source13:18
apevecactually, we decided to create rdo-tools :)13:19
dmsimardapevec: extras ?13:19
pabelangerEmilienM: would that fix your issues, if we mirrored?13:19
apevecthere was thread13:19
dmsimardapevec: the shade one ?13:19
EmilienMpabelanger: probably13:19
apevecextras has bad sound to it :)13:19
dmsimardapevec: I didn't follow too much13:19
apevecdmsimard, yeah that one13:19
pabelangerEmilienM: ok, let me see about pushing up a patch13:20
apevecpabelanger, please CC me13:20
apevecto make sure right repos are covered13:21
EmilienMme too13:21
pabelangerit will likely just be openstack lastest release, which we bump every cycle13:22
apevecEmilienM, but, it might be just networking issues on rh1 cloud side13:22
*** jprovazn has joined #rdo13:22
apevecpabelanger, yes,
derekhEmilienM: I don't think the mirror is the problem, we are loosing packets somewhere and this is causing problems with DNS (as its UDP)13:22
derekhapevec: pabelanger ^13:22
apevecthis is what trunk master jobs use13:22
apevecderekh, so it is errno 14 indeed13:23
aderyuginapevec: hi, can you please take a look at this patch is required to unlock upstream murano testing13:23
pabelangerderekh: that could be the issues too.  I know the networking setup for tripleo-test-cloud-rh1 is not the best13:23
pabelangerwe really should be pushing for an unfiltered pipe to the internet13:24
dmsimardaderyugin: hey, you know, I've been looking at those tempest results of yours yesterday13:24
*** jpena|lunch is now known as jpena13:24
*** amuller has joined #rdo13:24
pabelangerand not be sitting behind a linux box13:24
*** mlammon has joined #rdo13:24
derekhapevec: pabelanger EmilienM e.g. I can reproduce this pretty easily
dmsimardaderyugin: I find it odd that tempest is trying to fetch files directly in /usr/lib/python2.7/site-packages/* ?13:24
derekhline 7 ^ dns request, no reponse, 5 second time and retry13:24
apevecaderyugin, why is that folder not installed by setuptools ?13:25
dmsimardaderyugin: shouldn't "configuration" files at worst be shipped to /usr/share ?13:25
dmellado+1 to dmsimard and apevec comments13:25
pabelangerderekh: which node are you running that on?13:25
dmsimardaderyugin: I'm referring to
dmsimarddmellado: +1 to what ?13:26
*** kgiusti has joined #rdo13:26
dmsimarddmellado: oh nm13:26
derekhpabelanger: an instance on the cloud, talking to named on the bastion13:26
dmelladodmsimard: did you get it? xD13:26
dmelladoI thought it was obvious, but maybe it wasn't13:26
dmsimarddmellado: thought you were referring to some old discussion13:26
dmsimardnot enough caffeine yet13:26
dmelladodmsimard: no worries, just get a double-shot coffee xD13:26
pabelangerderekh: moving to #tripleo13:27
derekhpabelanger: ok13:27
aderyugindmsimard: murano-tempest-plugin was designed to run on devstack and murano developers doesn't plan to rewrite it13:28
apevecworked-in-devstack(TM) !13:28
*** mflobo has quit IRC13:28
apevecwe actually have a label in RDO CI for that13:28
dmsimardaderyugin: and devstack configures tempest in such a way that it tries to fetch stuff from /usr/lib/python2.7/* ? That's nuts13:29
dmelladoaderyugin: the tempest plugin should work wherever it's run13:29
EmilienMapevec: do we? lol.13:29
apevecWD - worked in devstack13:29
dmelladoit's just an entry point and some config options, what's the issue?13:29
dmsimardaderyugin: I'm not shooting the messenger, I know it's not your fault but that's just silly13:29
dmelladoapevec: I even have that t-shirt13:29
apevecEmilienM, top of
dmsimardaderyugin: like, tempest is trying to read configuration directly out from the module source. The config should be in /etc/ or /usr/share/13:30
*** sdake has joined #rdo13:30
dmsimardThis is just bad practice13:30
*** jlibosva has quit IRC13:30
dmsimardand should be filed as a bug13:30
*** gszasz has quit IRC13:31
EmilienMWD is the best!13:31
*** jjoyce has joined #rdo13:31
*** pradk has quit IRC13:32
*** pradk has joined #rdo13:32
dmsimardaderyugin: I'm not going to block the change, you do what you gotta do but it's not awesome13:33
dmelladoaderyugin: dmsimard13:34
dmelladocan't this be changed to some proper location13:34
*** jlibosva has joined #rdo13:34
dmelladoor maybe put it into its conf?13:34
dmsimardit opens the file with write privileges ?13:34
dmsimardwho thought that was a good idea13:35
apevecthis will of course fail when runnin as non-root ?13:35
apevecwith rpm install13:36
dmelladoapevec: that's my point too13:36
aderyugindmellado: python script you've posted uses tmp directory to write manifest, I've fixed it some days ago by myself13:37
dmsimardaderyugin: ok so the plugin wants to copy that directory to a temporary folder and then do the config in that temporary folder and use that ?13:38
*** jschluet has joined #rdo13:38
rdogerritJavier Peña created rdoinfo: Pin oslo.db in newton-uc to 4.13.3
dmelladoaderyugin: dmsimard my concern is for the rpm13:38
*** Amii is now known as Amita13:39
aderyugindmsimard: yes, now it work in this way13:39
dmsimarddmellado: so it doesn't write in /usr/lib after all13:40
*** jmelvin has joined #rdo13:40
dmsimardit needs read to copy to a tmpdir13:41
dmsimardand then uses that tmpdir13:41
*** jschlueter has quit IRC13:41
dmsimardthe patch adds a directory that was missing apparently13:41
*** jschluet is now known as jschlueter|afk13:41
*** jschlueter|afk is now known as jschlueter13:41
dmelladodmsimard: that's my guess too13:41
dmelladofrom the logs13:41
dmelladothe error was from that part13:41
dmellado2016-09-06 16:33:02.909491 |       File "/usr/lib/python2.7/site-packages/murano_tempest_tests/", line 56, in compose_package13:42
dmellado2016-09-06 16:33:02.909509 |         with open(manifest, 'w') as f:13:42
dmellado2016-09-06 16:33:02.909547 |     IOError: [Errno 13] Permission denied: '/usr/lib/python2.7/site-packages/murano_tempest_tests/extras/MockApp/manifest.yaml'13:42
dmelladosorry for the paste13:42
dmelladodmsimard: could you check if that patch is applied on the rpm build?13:42
aderyugindmellado: fix for tests was merged couple of days ago13:43
*** rhallisey has quit IRC13:43
*** rhallisey has joined #rdo13:44
dmelladoaderyugin: then my guess is that it didn't make it to the build13:44
dmelladojpena: dmsimard can we check that the rpm is based on?13:45
* jpena checks13:45
jpenadmellado: it's in the current/ repo, which job was failing?13:47
*** Son_Goku has quit IRC13:50
dmsimardjpena: it's not a new issue, murano testing never worked -- they're just trying to get it to work from
jpenasince (merged ~20 hours ago), the puppet-openstack-integration CI is using a repo with the patch13:52
jpenathe minimum version required is openstack-murano-api-3.0.0-0.20160908151553.1833842.el7.centos.noarch.rpm, note 1833842 is the short hash for the test fix13:52
dmelladojpena: dmsimard could we retrigger that job and check?13:52
*** chkumar|out is now known as chandankumar13:53
* chandankumar back13:53
jpenadmellado: I see is the review that enables murano in puppet-openstack-integration, right?13:54
jruzickatime for some czech tekno - see you next week, #rdo o/13:54
EmilienMsileht: can I blame you today? :)13:55
silehtEmilienM, yes and I will bother gordc, next13:55
EmilienMnot sure why yet but ceilometer now requires gnocchiclient13:56
EmilienMmaybe it's in requirements somewhere13:56
EmilienMit's not in
silehtEmilienM, it's optional13:56
dmelladojruzicka: IIUC yes13:57
silehtEmilienM, it's when you use gnocchi as backend13:57
* jpena clicks recheck13:57
EmilienMsileht: yeah but RDO uses requirements to fin ddependencies13:57
silehtEmilienM, yeah perhaps we should add it in requirement13:57
EmilienMsileht: yes IMHO you should13:57
EmilienMit doesn't hurt13:57
jruzickathe recheck I mean13:57
EmilienMsileht: let me propose it13:57
dmsimardEmilienM, sileht: if it's an optional dependency, the user/intaller should install it13:58
dmsimardRDO typically doesn't add requirements on optional dependencies13:58
dmsimardjust like we (exaggerating) don't install gluster, ceph and swift when installing glance13:58
silehtEmilienM, Gnocchi is the default now, so having it in requirement make sense13:58
dmsimardsileht, EmilienM: ok, if you have a patch to add it to reqs.txt we'll bundle it.13:59
EmilienMdmsimard: i'm working on it13:59
chandankumardmsimard: regarding this , tempest is failed to build on trunk.13:59
chandankumarnow i am seeing DLRN has passed13:59
dmsimardchandankumar: WIP14:00
chandankumardmsimard: ok14:00
dmsimardchandankumar: it's in mitaka, not master14:00
dmsimardchandankumar: fixed by
chandankumardmsimard: yes, that i thought.14:00
silehtdmsimard, EmilienM,
EmilienMsileht: why the package is not python-gnocchiclient?14:00
silehtEmilienM, for fun14:01
*** fragatina has joined #rdo14:01
pabelangerdmsimard: number80: apevec: Regarding 3rdparty CI, currently smtp is support by ZUUL. If we want to send failures some other place, we need to write some sort of job to handle that. Or improve zuul14:01
EmilienMsileht: no constraint?14:02
dmsimardpabelanger: did you do a typo in that ?14:02
silehtEmilienM, just added14:02
number80pabelanger: smtp is fine w/ me14:03
*** pkovar has quit IRC14:03
pabelangerdmsimard: typo?14:03
*** pnavarro has quit IRC14:04
EmilienMsileht: that's why I wanted to do it :P14:04
EmilienMsileht: thanks !14:04
dmsimardpabelanger: I'm not sure I understand the sentence, it seems like you're saying Zuul doesn't support SMTP but you're also saying it's supported14:05
dmsimardpabelanger: probably just me that's tired though14:05
*** nyechiel has quit IRC14:05
silehtEmilienM, did you connect you brain directly to the computer ?14:05
EmilienMsileht: sometimes :P14:05
*** fragatina has quit IRC14:06
pabelangerdmsimard: Zuul support both gerrit and smtp for success / failures messages.  If we want something else, we need to write something to handle that14:06
EmilienMsileht: I would also have written a commit message14:06
number80sileht: the other times, he just play the mechanical turk and put himself inside the computer14:06
*** akshai_ has joined #rdo14:08
rdogerritchkumar246 created openstack/tempest-distgit: Added python-gabbi and python-aodh-tests as Requires for openstack-tempest
*** myoung has joined #rdo14:09
chandankumardmsimard: does python-designate-tests-tempest, python-horizon-tests-tempest and python-watcher-tests-tempest packages are available in Mitaka?14:11
dmsimardchandankumar: no14:11
*** akshai has quit IRC14:11
dmsimardthey're new packages from newton afaik14:11
chandankumardmsimard: then sending patch on ftbfs.14:11
dmsimardchandankumar: no patch to be done14:12
dmsimardchandankumar: it will be fixed by the rdoinfo review I sent you14:12
dmsimardchandankumar: dlrn is pulling rpm-master instead of rpm-mitaka for mitaka because rdoinfo has master source branch for mitaka release14:12
*** myoung is now known as myoung|bbl14:12
chandankumardmsimard: ack, got that!14:13
chandankumarnumber80: time to unblock this review
chandankumardmsimard: please update this review when you are free!14:16
dmsimardchandankumar: yeah it's on my todo14:16
chandankumaronly heat tempest subpackage is pending,14:17
number80I'll remove -W when it will be updated, -W is to prevent accidental merge14:17
*** satya4ever has quit IRC14:18
jpenachandankumar, we've created a circular dependency around tempest and tempest plugins, see
*** pkovar has joined #rdo14:22
dmsimardjpena: hey can we add https to delorean.repo ?
jpenadmsimard: sure why not14:24
chandankumarjpena: will fix a new review14:24
chandankumarThanks for the pointer.14:24
*** jprovazn has quit IRC14:24
jpenachandankumar: I'm checking how many packages are involved in the circular dep, will let you know14:25
chandankumarjpena: sure!14:25
chandankumarjpena: currently we have horizon and designate only i think.14:25
number80jpena: not really, the problem is that DLRN-rpmbuild does a mock --install14:27
number80so when you build tempest, you need bootstrap to disable building --all14:28
number80chandankumar: ^14:28
rdobot[sensu] NEW: - check-delorean-newton-head-current @ |#| Build failure on centos7-master-head/current: ironic-inspector, python-tripleoclient:
*** pbandark has quit IRC14:30
*** mburned is now known as mburned_out14:30
EmilienMsileht: can you check if was done in RDO?14:30
EmilienMcheck if latest consistent URL contains the right version14:30
*** shaunm has quit IRC14:31
chandankumarnumber80: so on bootstraping, only python-tempest, openstack-tempest and python-tempesta-tests will got installed by mock, how can i make sure that all requires got installed for openstack-tempest-all subpackage?14:31
silehtEmilienM, the failing job is a newton one ?14:31
EmilienMsileht: yes14:31
EmilienMsileht: I guess we need to update requirements14:31
EmilienMin rdo14:32
EmilienMlet me look14:32
jpenachandankumar: the idea behind bootstrap is that you put the whole -all subpackage inside a %if 0%{?repo_bootstrap} == 0. It will first be built without -all, but then marked to be retried in the DLRN db, after all the dependant packages have been successfully built14:32
number80chandankumar: just make -all package building only when bootstrap is disabling (much ike we do with python3)14:32
dmsimardEmilienM: I'll update upper-constraints14:33
EmilienMdmsimard: ah thanks14:33
EmilienMdmsimard: in rdo?14:33
chandankumarnumber80: got that!14:33
dmsimardEmilienM: er, ceilometer is still 2.5.0 in upper-constraints
dmsimardEmilienM: so nothing in RDO or upstream uses a release beyond that14:35
dmsimardneed to bump u-c to 2.6.1 if we want that14:35
silehtEmilienM, I wonder why upstream upper-contrainst have not been updated automatically14:35
dmsimardsileht: there is a review for it
dmsimardrequirements freeze14:36
dmsimardyou need to apply for FFE if that is required for newton14:36
silehtdmsimard, I wonder upper-constraint is not 2.6.014:37
silehtI can understand for 2.6.114:37
EmilienMwe have a situation here14:38
dmsimardsileht: same thing for 2.6.0 :P
dmsimard-2 req freeze14:38
dmsimardreq freeze Aug 29-0214:38
EmilienMcan we manually promote it in RDO?14:40
EmilienMit's really a bug IMHO and we can't release newton without this ceilometerclient release14:40
EmilienMit's blocking aodh/gnocchi stuffs14:40
silehtI'm tire about the requirement stuffs14:41
*** Son_Goku has joined #rdo14:41
dmsimardpradk, sileht: gnocchiclient is a dependency of which ceilometer subpackage ?14:42
dmsimardI'm writing the review now14:42
dmsimardEmilienM: the first step would be to apply for FFE14:42
*** hynekm has quit IRC14:42
silehtdmsimard, collector and the package that have the dbsync binary14:42
dmsimardEmilienM: we can see what to do if the FFE is denied but we shouldn't override upper-constraints14:42
EmilienMsileht: can you take care of the FFE request?14:43
pradkdmsimard,whats the issue?14:43
pradkreading up14:43
*** manous has quit IRC14:43
dmsimardpradk: gnocchiclient was a added as a runtime dependency when running with gnocchi but gnocchiclient wasn't in reqs.txt so we missed it14:44
dmsimardpradk: it's being added to reqs.txt in so we'll add it in RDO too.14:44
rdogerritDavid Moreau Simard created openstack/ceilometer-distgit: Add python-gnocchiclient to ceilometer-common as a runtime dependency
dmsimardsileht, EmilienM, pradk ^14:44
EmilienMalso ceilometerclient needs a bump in reqs/uc14:44
EmilienMwhich would require FFE14:45
pradkdmsimard, k s we need to add new requires in ceilo spec14:45
*** READ10 has joined #rdo14:45
pradkcool you got it already14:45
*** READ10 has quit IRC14:45
*** READ10 has joined #rdo14:45
*** mcornea has quit IRC14:46
dmsimardpradk: yeah14:46
dmsimardpradk: the main problem now is that it seems we need ceilometerclient 2.6.1 but both 2.6.0 and 2.6.1 have been denied in upstream upper-constraints bump due to requirements freeze14:47
dmsimardSomeone needs to apply for FFE quickly14:47
EmilienMsileht, pradk^ your the best persons to do it14:47
pradkdmsimard, added a small comment, lets use atleast 2.2.014:48
pradkdmsimard, so we need to aply that to ?14:49
pradkyea currently its -214:49
rdogerritDavid Moreau Simard proposed openstack/ceilometer-distgit: Add python-gnocchiclient to ceilometer-common as a runtime dependency
EmilienMpradk: yes14:50
dmsimardpradk: I'm not familiar with the FFE process :(14:50
EmilienMand really, I would suggest to bump it in RDO anyway14:50
pradksend an email with FFE sub?14:51
EmilienMpradk: let's talk about it on telemetry14:51
*** dtrainor has quit IRC14:51
pradksileht, you want to take care of the FFE? or should i?14:51
*** dtrainor has joined #rdo14:51
pradkdmsimard, thx +214:53
*** jcoufal has quit IRC14:53
silehtpradk, goahead, this process makes me sick14:55
jpenadmellado dmsimard murano tests in P-O-I still fail, but now for an unrelated reason ->
dmelladooh, lol14:56
dmelladolet me check14:56
*** myoung|bbl is now known as myoung14:56
*** jcoufal has joined #rdo14:56
pradksileht, looks like jd is searching for his 10m pole.. lets see ;)14:57
rdobot[sensu] NEW: - check-delorean-mitaka-current @ |#| Build failure on centos7-mitaka/current: tempest.git:
dmsimardpradk: I think the PTLs are usually the ones requesting FFEs ?14:57
dmsimardI could be wrong14:57
dmsimardor at least they have to be cleared by the plt14:57
*** pkovar has quit IRC14:58
pradkdmsimard, donno, i assume anyone should be able to do it.. its just an email and a commit referencing it14:58
number80not necessarily a PTL, at least a core14:58
silehtptl should just put +1 on the review14:58
rdogerritJavier Peña created openstack/openstackclient-distgit: Remove Tempest as BR, add openstacksdk
dmsimardah ok14:59
*** kenhui has joined #rdo15:00
dmsimarddmellado: so tempest is branchless but still cuts releases15:00
*** jobewan has joined #rdo15:00
dmsimarddmellado: could we not just associate a particular release to an openstack version ?15:00
dmsimarddmellado: not unlike what we do with upper-constraints right now or some other projects like telemetry15:00
dmelladodmsimard: that's more or less what we're doing so far with tags15:00
dmelladofor now 12.2 => newton15:01
*** abehl has quit IRC15:01
dmelladobut I might bump that up15:01
dmsimardyeah, that's downstream15:01
*** smeyer has quit IRC15:01
dmelladoafter the mid-cycle15:01
dmsimardbut I mean once we move to upstream15:01
dmelladowe could just do the same and it was my idea15:01
dmsimardchandankumar: still around ?15:01
chandankumardmsimard: yes15:01
rlandymandre: ping re: testday validations15:01
dmsimardchandankumar: I'm trying to understand the different between python-tempest and openstack-tempest15:02
*** pilasguru has joined #rdo15:02
dmsimardchandankumar: to add the right package to designate plugin15:02
*** pkovar has joined #rdo15:02
chandankumardmsimard: python-tempest is just a python library which can be used a python module without tempest executables15:02
dmsimardI guess the dependency should be python-tempest but it was openstack-tempest before
chandankumarso in designate if tempest is import in the code python-tempest will fill the gap15:03
chandankumaropenstack-tempest is contains tempest executables and magical tempest config tool which install python-tempest.15:03
aderyugindmellado: gates are finished, and now it's fails with another error: as I know it's related to missing class definition which is stored on a folder that I've added in my patch15:04
dmsimardchandankumar: so python-tempest should be ok, right ?15:04
mandrehey rlandy, what's up?15:04
chandankumardmsimard: yup!15:04
rlandymandre: so I thought I would take a short cut - I had a newton install from master and I wanted to install the validations package.15:05
rlandymandre: iow, I didn't install directly from but rather rdo master15:05
rdogerritDavid Moreau Simard proposed openstack/designate-tempest-plugin-distgit: Change openstack-tempest dependency to python-tempest
dmsimardchandankumar, number80, jpena ^15:06
*** alexismonville has quit IRC15:06
rlandymandre: ' 'No package openstack-tripleo-validations available'15:06
dmsimardchandankumar: you're so quick :)15:06
rlandymandre: was the package just included in the testday rpm?15:07
*** jaosorior has quit IRC15:07
chandankumarjpena: apevec number80 dmsimard EmilienM please have a look when you have time
dmelladoaderyugin: so then we're expecting another patch for that?15:07
rlandymandre: I see the
rlandyI could just clone the repo, but I guess that's counter to testing packaging15:07
*** pcaruana has quit IRC15:08
mandrerlandy: hmm, I'm not really familiar with how testdays are supposed to happen... all I can say is that the package is available in
*** ihrachys has quit IRC15:08
*** mburned_out is now known as mburned15:09
rlandymandre: ok - so I need to find out the diff from what I installed and rdo-trunk-master-tripleo15:09
*** rpioso has joined #rdo15:10
*** abregman has quit IRC15:10
mandrerlandy: maybe :) like I said, I'm not the expert here, idk which package repo we're supposed to test during testdays15:11
rlandymandre: I am asking on #oooq ... will get back to you15:12
dmsimardjpena: gerrit privs fixed for designate plugin, log out and back in, should be ok15:12
rlandyadarazs: pls see ^^ from mandre15:12
rlandythe package I need (validations) is in
rlandyI am installing with --release master flag15:13
rlandywhat am I missing?15:13
adarazsrlandy: so I'm not sure about what this repo is. but I suspect that it's based on the tripleo ci's current-tripleo pin15:14
rlandyadarazs: ok - if so, how far out of sync our the repos we use?15:15
adarazsrlandy: uh, I don't think I can untangle what that repo is, and what is the correlation between the different versions now :)15:16
adarazsrlandy: what we should use for the testday is
*** egafford has joined #rdo15:17
aderyugindmellado: yes, my patch  adds missing directory15:17
rlandyadarazs: yep - understood15:18
rlandymoving back to #oooq15:18
adarazsrlandy: which is based on this hash:
*** vaneldik has quit IRC15:18
jpenadmsimard: ack, thanks15:19
*** egafford has quit IRC15:21
apevecrlandy, what is  rdo-trunk-master-tripleo ?15:22
apevecah I see link15:22
*** mbound has quit IRC15:22
*** jcoufal has quit IRC15:22
*** egafford has joined #rdo15:22
rlandyapevec: I am trying to work this all out myself15:22
rlandyie: get the right install going15:22
apevecrlandy, it has validations
*** Son_Goku has quit IRC15:23
*** nyechiel has joined #rdo15:24
rlandyapevec: I had a master install already and I was hoping to be able to use it and just install validations over it - but I am thinking that hack is not really serving the testday purpose15:24
apevecif that's was older install, you need to update .repo15:25
*** mosulica has quit IRC15:26
apevecyou could just yum install to get the testday repo15:26
*** shaunm has joined #rdo15:29
chandankumardmsimard: jpena dmellado
chandankumarit was due to aodh15:29
dmelladoI was replying to the gerrit, exactly that15:29
dmelladowe'd need to agree on the usecase15:30
chandankumardmellado: this patch does not make sense, it was due to aodh15:32
dmelladochandankumar: then please abandon it ;)15:32
dmelladoI'm off for today, have a great weekend y'all!15:33
*** leanderthal is now known as leanderthal|afk15:35
*** jistr is now known as jistr|biab15:35
rdogerritMerged openstack/nova-distgit: add package for placement API service
rdogerritMerged openstack/designate-tempest-plugin-distgit: Change openstack-tempest dependency to python-tempest
rdogerritPradeep Kilambi created openstack/aodh-distgit: Rotate logs as they reach 10M size
*** jcoufal has joined #rdo15:37
rdogerritPradeep Kilambi created openstack/gnocchi-distgit: Rotate logs as they reach 10M size
*** nyechiel has quit IRC15:40
number80pradk logrotate patchset are easyfix, please review them or I'll just merge them tonight ^15:40
pradknumber80, i can review my own fix? :)15:41
*** karimb has quit IRC15:41
*** eaguilar has joined #rdo15:43
*** kenhui has quit IRC15:44
number80pradk: preferably, reviews should be done by peers :)15:45
pradknumber80, oh i thought you were asking me to review above .. i misread15:46
number80pradk: no, just giving a chance for others to weight in15:46
pradkgot it :)15:47
*** openstackgerrit has quit IRC15:49
*** tesseract- has quit IRC15:49
*** openstackgerrit has joined #rdo15:50
*** ChanServ sets mode: +v openstackgerrit15:50
number80dmsimard: for please wait patch to be merged upstream15:51
number80if for reason X, it gets refused, we end up carrying useless deps15:51
*** aufi has quit IRC15:52
number80FTBFS is ok, as we have patchset ready to merge15:55
derekhapevec: kbsingh as a mater of interest, the redirects we were discussing earlier, how long have they been going on?15:55
-openstackstatus- NOTICE: New setuptools release appears to have a circular import which is breaking many jobs - check for ImportError: cannot import name monkey.15:55
*** ChanServ changes topic to "New setuptools release appears to have a circular import which is breaking many jobs - check for ImportError: cannot import name monkey."15:55
*** jcoufal has quit IRC15:56
*** jcoufal has joined #rdo15:56
apevecderekh, since we moved to cdn which is long ago15:59
*** dtantsur is now known as dtantsur|afk15:59
apeveci.e. last few months (don't remember the date)15:59
apevecnot last few days15:59
derekhapevec: ok15:59
apevecjpena, dmsimard ^ do you remember when was cdn turned on?15:59
derekhapevec: it doesn't matter exactly, I was just wondering if it was this week16:00
*** karimb has joined #rdo16:00
jpenaapevec: it was during this cycle. However, I'm not sure when it started to be used fully in Tripleo CI. Remember all they did before was to fetch delorean.repo, which then made them use directly16:00
derekhapevec: I think i found the problem we were having , the conntrack table on our bastion host was filling up and as a result it was dropping packets16:00
apevecuh oh16:01
apevecthanks for tracking that down16:01
derekhapevec: I was wondering if the redirects suddenly caused it to start happen as it causes us to do twice as many dns requests16:01
derekhapevec: the table was pretty much full with UDP streams being tracked for DNS16:02
derekhapevec: reducing some timeouts seems to have helped a lot, will know for sure in a few hours when results start to come in16:02
zoli|wfhhave a good weekend16:04
*** zoli|wfh is now known as zoli|gone16:04
rdogerritjason guiditta proposed puppet/puppet-ovn-distgit: Add spec for puppet-ovn.
*** limao has joined #rdo16:08
rdogerritchkumar246 created openstack/tempest-distgit: Added repo_bootstrap macro in order to bootstrap to disable building --all subpackage in dlrn
chandankumarjpena: number80 not sure about proper commit message
chandankumarplease suggest16:09
*** akshai_ has quit IRC16:12
*** limao has quit IRC16:12
*** pbandark has joined #rdo16:13
chandankumarnumber80: apevec jpena dmsimard jruzicka rbowen EmilienM Good Night, Have a nice weekend, see you on monday :-)16:14
dmsimardapevec: what cdn turned on for what ?16:14
EmilienMchandankumar: see you!16:14
*** chandankumar has quit IRC16:16
dmsimardnumber80: were you the one that created the projects in gerrit for puppet-barbican ?16:18
number80dmsimard: yes16:18
dmsimardnumber80: did you create them sort of manually through sfmanager ?16:18
*** shaunm has quit IRC16:18
dmsimardnumber80: also, how do we have a "reverse" replication ?16:18
dmsimardfor the source package16:18
number80originally with RDO script and then I redid them by hand16:19
number80I'd like to know who did the other puppet btw, as it's a mess16:19
*** sreichar has joined #rdo16:19
dmsimardprolly sf guys ?16:19
*** jlibosva has quit IRC16:19
dmsimardI was hoping the process to be a little bit more clean, I need to create puppet-designate and puppet-magnum16:20
*** lmiccini has quit IRC16:21
*** florianf has quit IRC16:21
number80dmsimard: I used rdo-create.py16:21
*** pbandark has quit IRC16:21
*** zoli|gone is now known as zoli_gone-proxy16:21
jaygnumber80: speaking of 'other puppet', any thoughts on the upstream_version question I had in puppet-ovn distgit review?16:21
number80I couldn't use RDO creation wrapper, as there's no github replication and namespace is different16:22
*** jcoufal has quit IRC16:22
*** mbound has joined #rdo16:22
number80jayg: looking16:22
* jayg was under (perhaps mistaken) impression the puppet packaging stuff was pretty stable at this point, happy to help improve it if that is not true16:23
*** jcoufal has joined #rdo16:24
number80jayg: comment left16:24
number80jayg: openstack packaging is a collection of exceptions :)16:24
jaygnumber80: hmm, that doesn't really answer the question though (I will add it if you want) - why do we want this here, but not on the modules we already have?16:25
number80so automating is painful, as you always reach weird stuff in a dark corner16:25
*** akshai has joined #rdo16:25
jaygpuppet may be a rather large 'dark corner'  ;)16:25
*** jistr|biab is now known as jistr16:25
number80jayg: the right question is how they builds in DLRN16:26
* jayg thought he was helping with some small fixes before moving toward OS cli, but it is not looking that way atm...16:26
jaygnumber80: ok, I have no clue there16:26
number80DLRN inject upstream_version so that it can build using DLRN generated tarballs16:26
number80well, if it passes CI, let's just merge them for now16:27
number80but I fear that CBS import will bring new issues to the light16:27
dmsimardayoung, stevemar: Is there a recommended backend for memcache support in keystone (for authtoken) ? I guess from the list of backends available there doesn't seem to be one above the others
jaygI deliberately left it out on ovn to see what jenkins did, it passed both with and without upstream_version16:27
ayoungdmsimard, nope16:27
*** mbound has quit IRC16:28
ayoungdmsimard, no use of memcache for backends recommended16:28
dmsimardayoung: but yet in-process token caching is being taken out16:28
ayoungdmsimard, tryin to get to fernet, to remove the need for a backend16:28
jaygnumber80: so if the right thing is to add it, I can put that in my queue for next week, unless someone like jpena has the answer to why it even works without it16:28
dmsimardayoung: ah, makes sense.. no memcache requirement with fernet ?16:28
ayoungdmsimard, patches are landing as we speak16:28
*** sdake has quit IRC16:28
*** alexismonville has joined #rdo16:28
rdobot[sensu] NEW: - check-delorean-newton-head-current @ |#| Build failure on centos7-master-head/current: ironic-inspector, python-tripleoclient:
dmsimardayoung: is that going to be in for newton ?16:28
ayoungdmsimard, we hope!16:29
dmsimardmwhahaha: fyi ^16:29
jaygnumber80: also, for the package reviews, it seems to me I should use RDO-NEWTON-PUPPET instead of just RDO-NEWTON, what do you think?16:29
ayoungdmsimard,  that is essential, and in the gate16:30
jaygBZs I mean16:30
*** jbadiapa has quit IRC16:30
ayoungonve we are through that, we should be able to test it16:30
jpenajayg: it's a bit of a lucky shot. If %{version} is the same as %{upstream_version}, there will be no issues in the %setup phase. When the calculation of %{upstream_version} provides a different value, it fails16:30
dmsimardayoung: ack, so just double checking -- there should be no memcache for authtoken config when using fernet tokens ?16:31
jpenathat's why we always use %{upstream_version} in spec files for DLNR16:31
jaygjpena: so likely all the other puppet modules has been rabbit-foot lucky to date  ;)16:31
ayoungdmsimard, should not be necessary.16:31
dmsimardayoung: we are setting memcached everywhere right now in puppet-openstack-integration so we'll take that out16:31
jaygI can put it on my todo list to fix them16:31
dmsimardthere were deprecation notices around in-process cache in logs16:31
ayoungdmsimard, so caching and backend are two separate things16:31
jaygI'll update puppet-ovn to match barbican with upstream_version in the meantime16:32
jpenajayg: or the way we derive %{upstream_version} from the metadata.json file prevents the issue most of the time16:32
ayoungdmsimard, caching still can benefit from memcache, but not storing tokens in the memcache backend16:32
*** fragatina has joined #rdo16:32
ayoungmemcached is unreliable, as it might dump data16:32
ayoungnot usable as a token backend16:32
jaygjpena: ah, k - thx16:32
*** READ10 has quit IRC16:32
dmsimardayoung: Ok, I understand. So it would be a keystone cache backend but not used for tokens (so not in [keystone_authtoken])16:33
*** snecklifter has quit IRC16:33
mwhahahadmsimard: yea we don't use it for a token backend with fernet16:33
mwhahahathe issue i was describing was for token validation16:33
mwhahahathe python-memcache configuration hates it when a server is down (increases request time)16:33
rdogerritjason guiditta proposed puppet/puppet-ovn-distgit: Add spec for puppet-ovn.
*** athomas has quit IRC16:34
ayoungdmsimard, ugh16:34 I think what you said is correct, but would confust just about everyone when phrased that way16:35
ayoungwhen we say backend, we mean persistnat stor16:35
ayounga cache is the opposite of that16:35
dmsimardayoung: there's a [cache] section and a [keystone_authtoken] section16:35
ayoungit is explicitly not a persistant store16:35
ayoungand that is soemthging all together different16:35
dmsimardayoung: you can set memcache config in either but they are different16:35
ayoungin KEYSTONE SERVER16:36
ayoungtokens must be PERSISTED in SQL today16:36
ayoungKEYSTONE SERVER has a cache used once it has validated a token16:36
ayoungthat cache can be backed by memcache16:36
ayoungit means that the backend does not have to refetch all the data during a token validation16:37
ayoungthat is step one16:37
rdogerritJavier Peña created rdo-infra/puppet-dlrn: Remove priority from dependency repos
ayoungstep 2, when a user calls in to Nova, Glacne, or some other service, it groust through authtoken middleware16:37
ayoungand there it calls keystone.  The response from keystone can be cached, and again, yes, this cache can be backed by memcache16:38
*** dsneddon has joined #rdo16:38
*** sreichar has quit IRC16:38
ayoungdmsimard, is that clear?16:38
dmsimardyes, it is16:38
ayoungnow, with Fernet, the only change is that Fernet tokens do not require the SQL backend16:38
ayounginside the KEYSTONE SERVER16:38
dmsimardstep 1 is [cache] in keystone server and step 2 is [keystone_authtoken] in cluents16:38
ayoungand memcache for both of those is a valid approach16:39
dmsimardright, okay.16:39
dmsimardThanks :)16:39
*** shaunm has joined #rdo16:43
rdogerritJavier Peña created DLRN: Revert "Revert "Enable yum priorities plugin for mock builds""
*** alexismonville has quit IRC16:44
*** akrivoka has quit IRC16:44
*** lucasagomes is now known as lucas-afk16:44
rdogerritpabelanger proposed config: Update openstack-check pipeline triggers
pabelangerdmsimard: care to review?^16:45
dmsimardpabelanger: wanted to chat with you about that in fact16:45
pabelangerdmsimard: we'll need to restart zuul16:45
dmsimardpabelanger: my only concern, which is not really one, but anyway16:45
dmsimardpabelanger: is that we surely might, one day, want to roll third party CI on every patchset, not necessarily only once they have been +1'd16:46
dmsimardso it's a nit but was wondering if we should change the pipeline name to openstack-check-verified or something to that effect16:46
*** lmiccini has joined #rdo16:47
pabelangerdmsimard: Sure, we do that today. I was mostly concerned about trying to package broken patches from upstream16:47
dmsimardpabelanger: right, I'm all for that particular dlrn job to run only after +v from jenkins16:47
dmsimardpabelanger: but "openstack-check" (upstream) doesn't wait until +v to run, it runs on every patchset16:48
dmsimardit's really a nit on the pipeline name16:48
dmsimardbecause eventually we'd have openstack-check that'd really run on each patchset16:48
dmsimardversus only once they have +v16:48
pabelangerRight, so we can rename that now too16:48
pabelangeror just create a new pipeline16:48
pabelangerand move DLRN jobs into it16:48
pabelangerlets do that16:48
dmsimardI'd probably name it opentack-check-verified but I have no strong opinion16:49
dmsimarddon't need openstack-check until we need it16:49
*** READ10 has joined #rdo16:49
dmsimardno need to create "empty" pipelines for nothing, it causes additional load on zuul for no reason16:49
pabelangerdmsimard: it might be worth keeping openstack-check for now, until we have all the DLRN jobs working.  Shouldn't cause more load, just additional configuration16:50
pabelangerbut, your call :) Is your zuul after all16:50
dmsimardpabelanger: from what I've seen in Zuul logs, Zuul keeps up with every patchset of upstream gerrit as I see it now16:50
dmsimardpabelanger: ah, I guess it would do it anyway no matter the amount of pipelines16:50
dmsimardok, then, sure16:51
pabelangerlet me update that patch with the new pipeline16:51
*** kenhui has joined #rdo16:51
dmsimardfeel free to go ahead with -verified (and change the layout to use that instead)16:51
*** akrivoka has joined #rdo16:52
*** tosky has quit IRC16:52
*** jpich has quit IRC16:52
*** markd_ has quit IRC16:53
dmsimardtrown: are you there today ? been quiet :)16:53
trowndmsimard: ya, been trying to get something working for quickstart16:53
*** derekh has quit IRC16:54
dmsimardtrown: ok, it's not an emergency but I'll just need some eyes on a ci-config review. Trying to create a promote-master pipeline for puppet-openstack similar to the tripleo one16:54
trownhad it this morning, and now all of the sudden it is not working, and I don't get why :(16:54
dmsimardsetuptools ?16:54
number80setuptools what?16:54
trownnah, setuptools thing was a slight hiccup in my day, but I am working on consuming an overcloud-full.qcow2 image directly16:55
dmsimardnumber80: there was a setuptools release that had a circular dependency issue16:55
trownhad it mostly working this morning, but now the thing wont even boot16:55
dmsimardnumber80: it was fixed rather quickly but everything exploded16:55
trownya, friday.releases--16:55
number80dmsimard: ok, no package downgrade request, I'm fine then :)16:55
trowngetting some food16:56
*** trown is now known as trown|lunch16:56
dmsimardthat's a good idea16:56
dmsimardafk :)16:56
number80still not hungry for dinner16:57
rdobot[sensu] NEW: - check-delorean-mitaka-current @ |#| Build failure on centos7-mitaka/current: tempest.git:
rdogerritpabelanger proposed config: Create openstack-check-verified pipeline
*** ohamada has quit IRC16:58
pabelangermessed up something there16:59
*** ChanServ changes topic to "RDO: An OpenStack Distribution for CentOS/Fedora/RHEL | Test day today: | Community meeting is held here every Wednesday at 15:00UTC | M3 Test Day September 8, 9 - | Please read our IRC community guidelines: | Logged at""16:59
-openstackstatus- NOTICE: setuptools 27.1.2 addresses the circular import16:59
*** dneary has quit IRC16:59
rdogerritpabelanger proposed config: Create openstack-check-verified pipeline
rdogerritpabelanger proposed config: Create openstack-check-verified pipeline
*** degorenko is now known as _degorenko|afk17:01
pabelangerokay, who do I need to bribe to stop using Fast Forward Only strategy17:01
*** rodrigods has quit IRC17:03
*** dgurtner has quit IRC17:03
*** rodrigods has joined #rdo17:03
pabelangertristanC: fbo_: Is that exposed via configuration in git? or is that a deployment setting in SF? Using Merge If Necessary for gerrit17:03
pabelangershould be setup in I think17:03
*** markd_ has joined #rdo17:05
pabelangerdmsimard: where are the acl files for gerrit stored?17:06
dmsimardpabelanger: there are none, it's in the UI17:08
dmsimardwell, the project permissions17:08
pabelangerdmsimard: Hmm, you don't manage permissions via git?17:09
dmsimardI wish17:09
dmsimardpabelanger: isn't upstream like that too ?17:09
pabelangerall of our acls are in git17:09
dmsimardok, I'll check later.. picking up food17:10
pabelangerdmsimard: sure17:10
pabelangerI need to better understand how projects are imported into review.rdoproject.org17:10
jpenaweekend calls, enjoy!17:10
*** hrw has quit IRC17:10
*** jpena is now known as jpena|off17:11
*** albertom has quit IRC17:17
*** pkovar has quit IRC17:19
pabelangerdmsimard: okay, I think managesf does it. But project config files look to be hardcoded:
*** albertom has joined #rdo17:25
*** bnemec is now known as beekneemech17:28
number80pabelanger: why no fast-forward?17:29
* number80 no found of merge commits17:29
number80it's useless crap cluttering the history17:29
pabelangerright, but now means everybody needs to constantly rebase patches on every merge17:30
*** jcoufal has quit IRC17:30
*** flepied has quit IRC17:30
number80pabelanger: well, reviewer can do it on the submitter behalf17:30
pabelangernumber80: but that adds a human into the pipeline17:31
pabelangeralso, zuul is using git merge so you are using 2 different merge methods between git-review and gerrit. and gerrit and check pipeline17:32
honzanumber80: we're adding a few dependencies to the tripleo-ui project right now --- what steps should we take to update our rpms?17:33
pabelangernumber80: git cherry-pick would be nicer if you want to avoid ugly merge messages in history17:33
pabelangerthen the submitter doesn't need to rebase at all17:34
pabelangerfor a while we did git cherry-pick in openstack-infra because we wanted a clean history.  But at the end of the day, we just accepted 'Merge if Necessary' to avoid the addition overhead to keep it clean17:36
number80honza: we can keep using tripleo repo, or I can create you a separate repo in rdo-common github org17:37
number80apevec, jschlueter ^17:37
number80jjoyce: ^17:37
honzanumber80: ok --- after the deps are updated, should i create a patch to update the spec files?17:38
mnasertop result searching for "delorean repo" .. hah.17:38
mnaserthought people would find it entertaining, /me is back to work17:38
*** fragatina has quit IRC17:39
pabelangerdmsimard: we should consider using the tripleo-ui project for the next openstack-check project.17:39
number80Yep, though, I'd like to review new deps (no ticket required)17:39
*** rgogunskiy has joined #rdo17:39
*** pnavarro has joined #rdo17:42
*** fragatina has joined #rdo17:43
jschlueternumber80: ack17:45
*** alexismonville has joined #rdo17:46
*** fragatina has quit IRC17:48
*** rgogunskiy has quit IRC17:50
*** trown|lunch is now known as trown17:55
*** kenhui has quit IRC17:59
*** kenhui has joined #rdo18:00
*** pnavarro has quit IRC18:00
*** akshai has quit IRC18:05
*** kenhui1 has joined #rdo18:07
*** kenhui has quit IRC18:07
*** READ10 has quit IRC18:10
*** akshai has joined #rdo18:12
*** unclemarc has quit IRC18:13
*** unclemarc has joined #rdo18:15
mnaseris designatedashboard currently packaged by rdo (it doesn't seem to be, or is it part of another package?)18:18
*** unclemarc has quit IRC18:22
*** unclemarc has joined #rdo18:23
EmilienMI don't see it in
EmilienMmnaser: I would say no18:23
mnaserEmilienM, ok thanks ill look into how i could help contribute some of that..18:25
jschlueterhonza: what are the new packages/deps you are working on?18:28
rdobot[sensu] NEW: - check-delorean-newton-head-current @ |#| Build failure on centos7-master-head/current: ironic-inspector, python-tripleoclient:
*** vaneldik has joined #rdo18:29
mnaseris there a good *-ui package to base it off of?18:30
*** flepied has joined #rdo18:30
*** fragatina has joined #rdo18:34
*** jcoufal has joined #rdo18:34
*** fragatina has quit IRC18:34
*** fragatina has joined #rdo18:34
*** noslzzp has quit IRC18:36
*** Guest_95843 has joined #rdo18:37
Guest_95843Allah is doing18:37
Guest_95843sun is not doing Allah is doing18:38
Guest_95843moon is not doing Allah is doing18:38
Guest_95843stars are not doing Allah is doing18:39
jschlueterhonza: thanks18:39
Guest_95843planets are not doing Allah is doing18:39
jschlueterdmsimard: kick Guest_95843 ?18:40
*** ChanServ sets mode: +o number8018:40
*** Guest_95843 was kicked by number80 (Guest_95843)18:40
number80I wish I could do kamehameha wave, I would have sent him straight to the sun18:41
*** number80 sets mode: +o jschlueter18:41
*** aortega has quit IRC18:52
*** aortega has joined #rdo18:52
rdobot[sensu] NEW: - check-delorean-mitaka-current @ |#| Build failure on centos7-mitaka/current: tempest.git:
rdogerritrdo-trunk created openstack/openstackclient-distgit: python-openstackclient: failed to build 0b20335
*** danielbruno has quit IRC18:59
*** dpeacock has quit IRC19:02
EmilienMpradk: so what's the plan for ceilometerclient?19:05
pradkEmilienM, thought jd was handling the ffe .. no?19:06
*** mlammon has quit IRC19:09
*** mlammon1 has joined #rdo19:09
EmilienMpradk: where?19:09
EmilienMI don't see any activity about it19:09
EmilienMand it's like super critical for us to have it19:09
pradkEmilienM, k i'll submit a ffe and submit a requirements patch19:10
pradkso we need 2.6.1 ?19:10
EmilienMpradk: patch is here: h
EmilienMI haven't seen any request19:10
EmilienMand to be honest, I would also push a pin in RDO even if not upstream19:11
EmilienMI think it will help to have the telemetry scenario green in tripleo19:11
dmsimardmnaser: hey, we don't seem to have designate-dashboard, no:✓&query=designate19:11
EmilienMdmsimard: ^ see my thoughts19:11
dmsimardmnaser: it looks pretty dead from looking at commit history ?
EmilienMapevec: could we make an exception?19:13
EmilienMi'm going to propose it19:13
*** danielbruno has joined #rdo19:13
dmsimardmnaser: maybe manila-ui and magnum-ui are good starting points:
dmsimardEmilienM: I'm going to repeat that I would like to see the FFE process started19:14
dmsimardEmilienM: and that if it isn't resolved reasonably quickly upstream, we can consider what to do on our end19:14
pradkEmilienM, can you give me a good justification why we need this version .. compiling email19:14
dmsimardEmilienM: Doing anything in RDO before even trying upstream is not a good strategy19:14
EmilienMpradk: sileht gave it earlier on telemetry channel19:14
EmilienMdmsimard: right but it's not the first time we do it temporarly19:15
EmilienMdmsimard: I'm having crazy time to make tripleo scenarios stable19:15
EmilienMwe don't have to be so rigid if we know a release would help us19:15
EmilienManyway, let me find the reason again19:15
dmsimardEmilienM: please start the FFE process :P19:15
dmsimardI'm not saying we won't do it19:15
pradkEmilienM, i dont see it19:15
dmsimardI'm just saying we should do things in the right order19:15
*** dneary has joined #rdo19:15
EmilienMI found a backtrace in Aodh19:16
EmilienMand it was fixed by
dmsimardHandling exceptions by overriding things in RDO is just that, an exception, we shouldn't be overriding things whenever it's convenient to do so, it'd become unmanageable19:16
dmsimardpradk: please cc me in the ffe19:16
pradkwill do19:16
dmsimarddms @19:16
EmilienMpradk: let me 5 min to give more context19:17
*** mlammon1 has quit IRC19:17
EmilienMpradk: so tripleo fails to create an alarm:
EmilienMsomething we can see in aodh logs:
EmilienMand sileht told me it's fixed by
*** mlammon has joined #rdo19:20
*** karimb has quit IRC19:20
pradki'm sending the following19:21
EmilienMso it seems like aodh is sending wrong return code to gnocchi19:21
*** mlammon has quit IRC19:21
pradkEmilienM, so this is not fixed with aodhclient?19:21
*** mlammon has joined #rdo19:22
EmilienMyou can also see it here:
*** lmiccini has quit IRC19:22
*** mlammon has quit IRC19:22
EmilienMpradk: yes email is good19:22
*** eaguilar has quit IRC19:22
*** ohochman1 has joined #rdo19:23
EmilienMpradk: even if it doesn't solve our bug in CI, it will at least remove the horrible trace in logs19:23
EmilienMthe actual error that is making the job failing is here:
EmilienMClient-side error: Resource e68cef8f-50cc-43c7-8c99-a51e2366dc5f does not exist format_exception19:24
EmilienMe68.. is the alarm that is not found19:25
EmilienMI'm wondering if it's a Heat resource issue19:27
EmilienMwhere alarm is created too late19:27
*** xb_ has quit IRC19:27
*** xb_ has joined #rdo19:29
pradkEmilienM, sent19:29
pradkEmilienM, if you can +1 the requirements patch to show some support19:30
EmilienMpradk: thanks19:31
EmilienMpradk: I have to go for an errand now, I'll look the rest later19:32
EmilienMI need to find why the heat template fails to create an alarm19:33
dmsimardFebruary for Ocata
EmilienMI'm afk, have a great week end all19:36
*** mlammon has joined #rdo19:37
*** sdake has joined #rdo19:38
*** fultonj has quit IRC19:39
*** fultonj has joined #rdo19:41
*** sdake has quit IRC19:42
*** sdake has joined #rdo19:44
*** rbrady has quit IRC19:44
*** ccamacho has quit IRC19:49
*** saneax is now known as saneax-_-|AFK19:54
*** vaneldik has quit IRC19:57
*** mlammon has quit IRC20:08
rdogerrithguemar created openstack/keystonemiddleware-distgit: Sync requirements
*** rpioso has quit IRC20:14
*** kenhui1 has quit IRC20:16
*** jayg is now known as jayg|g0n320:18
*** mbound has joined #rdo20:20
*** hynekm has joined #rdo20:21
*** kgiusti has left #rdo20:21
*** aortega has quit IRC20:23
*** dciabrin has quit IRC20:26
rdobot[sensu] NEW: - check-delorean-newton-head-current @ |#| Build failure on centos7-master-head/current: ironic-inspector, python-openstackclient, python-tripleoclient:
*** pabelanger is now known as pall20:29
dmsimardOuch, fedora atomic is >500MB ?20:29
*** paragan has quit IRC20:30
*** kenhui has joined #rdo20:30
dmsimardThat's fairly large, no ?20:30
number80nope, though it's not minimized20:31
dmsimardCoreOS is <300MB20:31
number80the Fedora Cloud image is instead quite small < 200 Mo20:31
dmsimardAnd I'm talking about qcow220:32
*** pilasguru has quit IRC20:32
number80dmsimard: CoreOS has no tools, you need to download special container which is fedora-based ...20:32
*** pilasguru has joined #rdo20:33
*** hynekm has quit IRC20:33
*** jkilpatr has quit IRC20:33
*** lmiccini has joined #rdo20:39
*** alexismonville has quit IRC20:39
*** dciabrin has joined #rdo20:40
*** jcoufal has quit IRC20:43
*** Tenhi_ has joined #rdo20:46
*** albertom is now known as albertom_afk20:47
rdogerrithguemar proposed openstack/keystonemiddleware-distgit: Sync requirements
*** brault has quit IRC20:50
*** Tenhi_ has quit IRC20:53
*** jkilpatr has joined #rdo20:56
*** openstackstatus has quit IRC20:57
rdobot[sensu] NEW: - check-delorean-mitaka-current @ |#| Build failure on centos7-mitaka/current: tempest.git:
*** pall is now known as pabelanger21:00
*** openstackstatus has joined #rdo21:00
*** ChanServ sets mode: +v openstackstatus21:00
*** openstack_newbie has joined #rdo21:00
*** mlammon has joined #rdo21:02
openstack_newbieHello, I am trying to install the latest version of OpenStack Newton (I mean the master) on a CentOS VM using Packstack21:03
openstack_newbieI am referring this link -
*** jmelvin has quit IRC21:05
openstack_newbieThe final step - packstack --allinone keeps on failing21:05
openstack_newbiewith some prerequisite rpms are not signed errors.21:05
*** akrivoka has quit IRC21:06
*** unclemarc has quit IRC21:06
number80openstack_newbie: how did you configure repo?21:06
number80master packages are not signed21:06
openstack_newbieI configured it using -
openstack_newbieyum -y install
*** mlammon has quit IRC21:09
openstack_newbieIs there anyway I can ask packstack to not do sign checks for the packages ?21:09
number80openstack_newbie: it's something configured in your repo config, not by packstack21:10
openstack_newbieok, so I should change the .repo file to not do gpgchecks ?21:10
rdogerritMerged openstack/openstackclient-distgit: Remove Tempest as BR, add openstacksdk
number80actually the package provides correct repo files21:11
number80do you have any other RDO repo configured?21:11
*** iranzo has quit IRC21:11
number80maybe pasting logs to could help21:11
*** paragan has joined #rdo21:12
*** paragan has joined #rdo21:12
openstack_newbieI just have rod-release.repo and rdo-testing.repo in my /etc/yum.repos.d/21:14
openstack_newbieOk, I will paste the logs.21:15
openstackgerritIvan Chavero proposed openstack/packstack: Change ironic identity to authtoken
*** jeckersb_gone is now known as jeckersb21:16
*** mbound has quit IRC21:19
*** pilasguru has quit IRC21:20
openstackgerritMerged openstack/packstack: Use lbaasv2 instead of v1
*** trown is now known as trown|outtypewww21:21
number80this package is not supposed to be signed21:23
*** aortega has joined #rdo21:23
*** mburned is now known as mburned_out21:29
*** mwhahaha has quit IRC21:32
*** akshai has quit IRC21:32
openstack_newbiegot it21:32
openstack_newbiethanks for the pointer number8021:33
openstack_newbieI got past that problem21:33
openstack_newbieIn my rod-release.repo file, I set gpgcheck to 0 instead of 121:33
openstack_newbieand the installation completed successfully21:33
*** abregman has joined #rdo21:34
*** kenhui has quit IRC21:34
rdogerrithguemar created openstack/automaton-distgit: Use for Source0 URL
*** aortega has quit IRC21:36
*** openstack_newbie has quit IRC21:36
*** aortega has joined #rdo21:37
apevecopenstack_newbie, have you checked
apevecyou need to set CONFIG_ENABLE_RDO_TESTING=y21:41
apevecrdo-release for newton does not exist21:41
apevecdmsimard, number80 - wishlist season started! I need to give a quick (like in next 5min :) estimate how much resources is required by rdo infra in the rdocloud21:47
apeveccan we get some stats from ci.centos usage?21:47
dmsimardA ballpark figure would be to use two flavors, one for TripleO, one for other things21:48
dmsimardI suggest using the same standard as openstack-infra for "other things" specs21:48
dmsimardsec, finding it21:49
number80apevec: 10/200/200/6021:50
number80planning large for upstream stuff21:50
dmsimardapevec: 8vCPU, 8GB RAM, 80GB Disk21:51
dmsimardapevec: I would go with 30x openstack-infra standard21:51
number80dmsimard: they're not 3o compliant btw21:52
dmsimardnot oooq* compliant21:52
dmsimardbecause tripleo runs jobs upstream with multinode on those VMs21:52
dmsimardin openstack-infra gate21:52
*** mwhahaha has joined #rdo21:52
dmsimardpabelanger, EmilienM and slagle have been working on that for a while21:52
number80dmsimard: yeah, but for testing, I use single-node setups21:52
apevecok, so number80's est + 30* openstack-infra? :)21:52
dmsimardbkero will soon be running tripleo in review.rdo too21:53
apevecthen 100* ? :)21:53
dmsimardapevec: well, to be fair, I think with tripleo included I would bump that to 5021:53
dmsimardat the very least21:53
dmsimardIf trown|outtypewww or weshay_pto would be here, they'd be asking for 32GB, 8 vCPU, 80GB VMs probably21:54
number80that's really a Xmas wishist :D21:54
apevecok, I'll add comment my ask is w/o oooq and leave that to Wes21:54
dmsimardnumber80: there is a LOT of CI on ci.centos21:54
number80dmsimard: yeah, and I expect even more, but I'm glad that there are capacity upgrades planned later21:55
pabelangerit depends on what you want to do21:55
dmsimardapevec: for me it's 50*openstack-infra standard as per (this includes dlrn, weirdo, upstream tripleo)21:55
dmsimardapevec: I have no clue about oooq21:55
number803 AWS-sized regions21:56
dmsimardapevec: added to includes: any other related projects (rdoinfo, rdopkg, what have you)21:56
pabelangerIf this is for RDO, I would stick with 8vCPU, 8GB RAM, 80GB Disk21:56
dmsimardpabelanger: yes.21:56
pabelangerif it is tripleo-jobs, that is a different storey21:56
pabelangerand wouldn't want to rush that21:56
dmsimardpabelanger: how so, are you not running tripleo jobs upstream in infra gate now ?21:56
dmsimardthen where's the problem 621:57
pabelangerbut, they are not the same as what tripelo-test-cloud-rh1 does21:57
pabelangeralso, tripleo jobs time out after 3hrs 10mins21:57
pabelangerthat should not be the default21:57
dmsimardso they're actually failing in upstream gate is that what you are saying ?21:57
pabelangerso, I agree more resource are needed for tripleo21:57
pabelangerI just don't know what that is21:57
pabelangerdmsimard: no21:58
dmsimardapevec: ok so I have no clue about tripleo at all then21:58
pabelangerdmsimard: they only run a fraction of the tests21:58
number80apevec: simpler would be splitting our use-cases21:58
pabelangerdmsimard: EmilienM knows more then me21:58
dmsimardnumber80: +1, I think there are "competing" customers21:58
pabelangerfor unit tests / packagings, 8GB is more then enough21:58
pabelangerdeployment / integrated testing, that need to be reworked I think21:59
number80dmsimard: yeah, let's have them define their own needs21:59
pabelangermore resources == faster run times21:59
dmsimardpabelanger: I think that falls into the restriction to two nodes, yes ?21:59
dmsimardWould the jobs run better with 3 ?21:59
number80apevec: since jpena is not around, did you request for DLRN-sized instance (2 or 3) ?21:59
pabelangerdmsimard: I think we should get the specific parties into a channel and see what people are testing with.  I've seen reports of 32GB and some 6GB, just depends on the type of test I guess22:00
*** elmiko is now known as _elmiko22:00
number80I'd bump disk sizes if possible to limit disk cleanups22:00
dmsimardnumber80: jpena|off and I are working on a secret project to move dlrn package building to nodepool22:01
dmsimardnumber80: I included it in my *5022:01
dmsimardwell, working is a big word. We've had discussions around the idea22:02
apevecnumber80, for now it's only totals quota, for aortega's planning22:02
aortegaballpark.. conservative estimation :)22:02
*** amuller has quit IRC22:02
apevecyes, we're increasing factor to stay conservative22:03
dmsimardyes, I'm being generous with the number I provided22:03
dmsimardthat amount would change a lot if we move some of the RDO stuff upstream like we might want to do22:03
pabelangerhow many nodes do you want?22:03
apevecaortega, actually I'll put formula in you sheet ;)22:03
dmsimardthat amount considers we're keeping everything RDO related in our yard22:03
pabelangerI think that is going to drive the quote22:03
dmsimardif we move, for example, package building upstream, that changes a LOT where things will be hapenning22:04
pabelangerusing as a base would be a good first iteration22:04
dmsimardpabelanger: yes that's what I liniked22:04
pabelangerdmsimard: yup, so 100?22:04
apevecyes, that was my factor :)22:05
dmsimardI'm okay with 100, I was just being realistic with 5022:05
dmsimardIf you ask for 100 and get 50, eh22:05
dmsimardIf we ask 50 and we get 25 now that's pretty ordinary22:05
pabelangerhow big is each compute node?22:06
apevecdmsimard, you got it how this works22:06
dmsimardapevec: :)22:06
dmsimardRCIP dev cloud admin told me the other day I was starting to get their cloud busy with the jobs I've been adding in review.r.o :)22:07
apevecdmsimard, was that a friendly warning before they kick us out?22:08
dmsimardnot a warning, I don't think we're in trouble yet -- and anyway we can't really go over quota anyway22:08
dmsimardit's just that their cloud was fairly idle before we started throwing things at it22:08
apevecso was ci.centos :)22:09
apevecRDO, troublemakers22:09
number80that's the unofficial name :)22:12
*** aortega has quit IRC22:17
*** pilasguru has joined #rdo22:17
*** aortega has joined #rdo22:19
rdogerrithguemar created openstack/automaton-distgit: Add python3 subpackage
*** apevec has quit IRC22:22
*** pilasguru has quit IRC22:22
*** abregman has quit IRC22:28
rdobot[sensu] NEW: - check-delorean-newton-head-current @ |#| Build failure on centos7-master-head/current: ironic-inspector, python-openstackclient, python-tripleoclient:
rdogerritMerged rdoinfo: Set Tempest branches for newton-uc and mitaka
*** rlandy has quit IRC22:33
*** sdake has quit IRC22:40
*** null_ref has quit IRC22:45
rdogerritMerged openstack/keystonemiddleware-distgit: Sync requirements
rdobot[sensu] NEW: - check-delorean-mitaka-current @ |#| Build failure on centos7-mitaka/current: tempest.git:
tristanCpabelanger: acl are stored in each projects using the meta/config branch, that's a gerrit default.23:00
tristanCpabelanger: so yes you can manage acl through git. Though SF did setup the necessary settings so rdo user only use the web interface to add and remove from groups23:01
tristanCpabelanger: why do you need to remove the fast-forward only policy ?23:02
*** melwitt has left #rdo23:03
*** melwitt has joined #rdo23:04
*** jobewan has quit IRC23:11
*** Ryjedo has quit IRC23:16
*** pilasguru has joined #rdo23:19
*** limao has joined #rdo23:20
*** sdake has joined #rdo23:22
*** pilasguru has quit IRC23:23
*** limao has quit IRC23:24
*** pradk has quit IRC23:27
*** aortega has quit IRC23:36
pabelangertristanC: okay, so can you do code reviews on the acls? Or is it just via the web ui in gerrit? Still a little confused23:39
pabelangertristanC: I wanted to see why the default was fast-forward only, as I find it and awkward policy compared to merge if needed23:40
tristanCpabelanger: yes of course, just checkout meta/config, edit and git review, that's standard gerrit workflow23:42
tristanCpabelanger: well sf team find it awkward to have all the "merge commit" and prefer the fast-forward only... (cc fbo)23:43
tristanCpabelanger: it's probably just a recommendation and can likely be changed to merge if needed, but is there a technical need for such change ?23:45
pabelangertristanC: so, maybe git cherry-pick would be a good middle ground.23:45
pabelangertristanC: not technical, but more to cut down on the spam needed to rebase patches when something is merged23:45
pabelangertristanC: since we don't actually need to rebase code but commit history23:46
pabelangergit cherry-pick might be friendly23:46
tristanCpabelanger: i'll have to check with fbo to make sure it doesn't break the gating workflow23:47
pabelangerit shouldn't, we ran it for a while in openstack-infra.  A few years, but in the end we just moved to merge if needed23:47
tristanCi personnaly find it annoying to have so many empty "merge commit", it makes it hard to follow git history. It may double the number of commit and they are not filtered by default23:49
pabelangerYa, I would git cherry-pick would be better in that case. Since as a contributor I don't have to care about commit order.23:52
pabelangerbut if zuul fails to cherry-pick the patch, then a rebase will be needed23:52
pabelangerwhich is the same workflow as merge if needed23:52
tristanCpabelanger: that's a good point, fbo: what do you think about changing from "fast forard only" to "cherry-pick" in gerrit submit type ?23:54
pabelangerokay, I see refs/meta/config now. Only downside of that approach, is I would have to clone all projects to manage their configuration, opposed to storing the acl in project-config and manage_projects to the update23:54
pabelangertristanC: I have to run, for now. Might be back in a bit23:54
tristanCpabelanger: I believe acl are inherited from All-Projects.git, what change do you need to do ?23:55
tristanCpabelanger: sure, have a good evening sir!23:55

Generated by 2.14.0 by Marius Gedminas - find it at!