Monday, 2019-07-01

*** michael-beaver has quit IRC00:32
*** dchen has quit IRC00:35
*** dchen has joined #openstack-infra00:36
*** efried has joined #openstack-infra00:43
*** zhurong has joined #openstack-infra00:59
*** ricolin has joined #openstack-infra01:04
*** imacdonn has quit IRC01:10
*** imacdonn has joined #openstack-infra01:11
*** corvus has quit IRC01:12
*** corvus has joined #openstack-infra01:12
*** bobh has joined #openstack-infra01:17
*** bobh has quit IRC01:22
* ianw googles Bonaire01:27
ianwlooks great! :)01:27
*** yamamoto has joined #openstack-infra01:48
*** rfarr has quit IRC02:03
*** rfarr has joined #openstack-infra02:04
*** tinwood has quit IRC02:08
*** tinwood has joined #openstack-infra02:10
openstackgerritMerged opendev/system-config master: Add OVH GRA1 mirror  https://review.opendev.org/66754302:15
*** auristor has joined #openstack-infra02:43
*** yamamoto has quit IRC02:44
*** yamamoto has joined #openstack-infra02:44
*** bhavikdbavishi has joined #openstack-infra02:53
*** auristor has quit IRC02:55
*** bhavikdbavishi1 has joined #openstack-infra02:56
*** bhavikdbavishi has quit IRC02:58
*** bhavikdbavishi1 is now known as bhavikdbavishi02:58
*** auristor has joined #openstack-infra03:10
*** bhavikdbavishi has quit IRC03:10
openstackgerritIan Wienand proposed opendev/system-config master: letsencrypt-install-txt-record: skip disabled hosts  https://review.opendev.org/66834403:10
*** bhavikdbavishi has joined #openstack-infra03:11
auristorianw: 5.2-rc7, tagged yesterday, includes the kafs fixes in case you might want to update the kernel on iad03:24
ianwauristor: thanks, yep that was what i was going to roll out to the new gra1 server before i found a small issue getting certificates for it above ^^ :)03:27
auristorsorry, I've been offline all weekend03:28
ianwme too :)03:28
ianwi don't think we're having any issues with kafs as such.  we can investigate the offline messages but they don't really seem to cause an issue i can see03:28
ianwthe hard crashes seem to be filecache related, although it's survived all weekend (though it's much quieter)03:29
auristorthe offline messages are the result of the volumes being taken offline for the release.03:29
auristorI started to review the retry logic with dhowells on Friday but we ran out of time.03:29
auristorThe reason I'm interested in reviewing the audit logs from the fileservers is that I'm trying to figure out of the RO on afs01 is being taken offline during the release or if its just the RO on afs02 that is offline.03:31
auristorThat will tell us whether kafs is failing to failover to afs01 from afs02.03:32
auristorIt might be a case of timing.   afs01 takes the RO offline to update its clone from the RW.  kafs sees that and marks the volume instance offline.  It then switches to afs02 but by the time it does, the RO on afs02 has been taken offline.03:33
auristorI will continue to review the retry logic this week03:36
ianwauristor: so basically, the same way i set the -cb value, just set auditlog to a file for dafileserver?  we determined with that spreadsheet that we should have enough space to run for a few days without issue03:38
auristorgreat03:38
*** psachin has joined #openstack-infra03:38
auristor-auditlog <log_path> -audit-interface file03:42
ianwok i've done afs01.dfw03:46
*** rfarr has quit IRC03:47
*** rfarr has joined #openstack-infra03:47
auristoris there logging to indicate when the vos release process begins and ends?03:49
ianwauristor: yep, it varies depending on the type of mirroring (rsyncing, using reprepro, etc) but all the cron jobs log their output.  unfortunately not publicly logged ATM (i'm working on that ... no reason they can't be, just not done ATM)03:52
auristorthats ok.  as long as you have access to it03:52
ianwok, afs01/02.dfw & afs01.ord all now have audit logging.  i will just update our etherpad with the details & how to turn it off, in case i get hit by a bus :)03:54
*** ramishra has joined #openstack-infra03:57
ianw#status log AFS auditlog enabled for afs01/02.dfw and afs01.ord AFS servers.  logging to /opt/dafileserver.audit.log.  notes, including details on how to disable again @ https://etherpad.openstack.org/p/opendev-mirror-afs03:58
openstackstatusianw: finished logging03:58
*** bhavikdbavishi1 has joined #openstack-infra04:02
*** bhavikdbavishi has quit IRC04:04
*** bhavikdbavishi1 is now known as bhavikdbavishi04:04
*** yamamoto has quit IRC04:08
*** yamamoto has joined #openstack-infra04:10
*** kjackal has joined #openstack-infra04:11
*** virendra-sharma has joined #openstack-infra04:12
openstackgerritMerged opendev/system-config master: letsencrypt-install-txt-record: skip disabled hosts  https://review.opendev.org/66834404:24
openstackgerritIan Wienand proposed opendev/system-config master: Experimental kafs support  https://review.opendev.org/62397404:25
openstackgerritIan Wienand proposed opendev/system-config master: ubuntu-kernel: role to use Ubuntu mainline kernels  https://review.opendev.org/66505704:25
AJaegerconfig-core, please review https://review.opendev.org/#/c/667900/ and https://review.opendev.org/667437 and https://review.opendev.org/66804504:33
*** ykarel has joined #openstack-infra04:42
*** yamamoto has quit IRC04:43
openstackgerritMerged openstack/project-config master: Fix ACLs for retired ansible-role-tripleo-cookiecutter  https://review.opendev.org/66790004:50
openstackgerritMerged openstack/project-config master: Removed unused ACL  https://review.opendev.org/66743704:50
*** ricolin_ has joined #openstack-infra05:03
*** ricolin has quit IRC05:06
*** yamamoto has joined #openstack-infra05:09
*** ykarel_ has joined #openstack-infra05:09
*** ykarel has quit IRC05:12
*** bradm has quit IRC05:13
*** udesale has joined #openstack-infra05:18
*** ykarel_ is now known as ykarel|meeting05:20
*** sshnaidm_ is now known as sshnaidm|ruck05:27
*** kjackal has quit IRC05:27
*** ociuhandu has joined #openstack-infra05:40
*** ykarel|meeting is now known as ykarel05:41
*** jbadiapa has joined #openstack-infra05:45
*** rpittau|mtg is now known as rpittau|afk05:49
*** virendra-sharma has quit IRC05:49
*** ociuhandu has quit IRC05:53
*** yamamoto has quit IRC05:53
*** ociuhandu has joined #openstack-infra05:54
*** ociuhandu has quit IRC05:57
*** ociuhandu has joined #openstack-infra05:58
*** udesale has quit IRC05:59
*** kjackal has joined #openstack-infra06:04
*** snapiri has joined #openstack-infra06:08
*** ricolin_ has quit IRC06:09
*** ricolin has joined #openstack-infra06:09
AJaegerthanks, ianw !06:10
*** AJaeger has quit IRC06:11
*** AJaeger has joined #openstack-infra06:13
*** dpawlik has joined #openstack-infra06:15
*** pgaxatte has joined #openstack-infra06:27
*** yamamoto has joined #openstack-infra06:31
*** pgaxatte has quit IRC06:31
*** pgaxatte has joined #openstack-infra06:32
*** ociuhandu has quit IRC06:36
*** yamamoto has quit IRC06:41
*** slaweq has joined #openstack-infra06:48
*** ccamacho has joined #openstack-infra06:50
*** ccamacho has quit IRC06:50
*** ccamacho has joined #openstack-infra06:50
*** rcernin has quit IRC06:55
openstackgerritIan Wienand proposed openstack/project-config master: Switch ovh/GRA1 to opendev.org mirror  https://review.opendev.org/66836106:57
ianwfungi / anyone: ^ a short sync update; nothing that is not also noted in https://etherpad.openstack.org/p/opendev-mirror-afs06:58
ianw1) note afs servers have audit logging right now.  might be easier to work with auristor in daylight hours on any required extracts06:59
ianw2) mirror.gra1.ovh.opendev.org should be operational and configured basically exactly the same as iad.rax.  It is logging to 6667 (iad.rax 6666) on our netconsole capture host06:59
ianwit is running upstream 5.2-rc7 kernel, which should have the afs fixes in it.  iad.rax still running our hand-built afs-next kernel07:00
*** ginopc has joined #openstack-infra07:01
ianwclarkb: i'd say for new hosts setup as opendev.org mirror using bionic but with openafs (1.8.3 from our ppa).  that is all in system-config -- so basically just copy out mirror.rax.dfw.opendev.org07:02
*** xek has joined #openstack-infra07:05
AJaegerianw: btw. what about https://review.opendev.org/663854 - do we want that? Then it needs a rebase...07:08
*** tesseract has joined #openstack-infra07:10
ianwAJaeger: ah, yeah, we could switch in rax.ord -- that's openafs 1.8.3 based like dfw.rax07:11
*** iurygregory has joined #openstack-infra07:15
AJaegerwant to base it on top of 668361? Your decision on what's the best way forward...07:19
*** yamamoto has joined #openstack-infra07:19
*** lucasagomes has joined #openstack-infra07:19
*** tosky has joined #openstack-infra07:25
ianwtrying to remember now where i got to with ord server07:26
openstackgerritMaksim Malchuk proposed openstack/diskimage-builder master: Cleanup: remove useless statement  https://review.opendev.org/66837207:30
*** witek has joined #openstack-infra07:40
*** pcaruana has joined #openstack-infra07:43
*** rpittau|afk is now known as rpittau07:44
*** ykarel is now known as ykarel|lunch07:46
openstackgerritIan Wienand proposed openstack/project-config master: Switch RAX ORD mirrors to new opendev.org mirrors  https://review.opendev.org/66385407:50
openstackgerritIan Wienand proposed openstack/project-config master: Switch ovh/GRA1 to opendev.org mirror  https://review.opendev.org/66836107:50
ianwAJeager / clarkb: ^ I think that's the way forward ... dfw has been running with openafs 1.8.3 without issue AFAIK.  so we can switch in ORD too with that; just leave our kafs testing on the two hosts (one local-ish, one remote-ish)07:51
*** ralonsoh has joined #openstack-infra07:52
AJaegerianw: Thanks, LGTM for that change ;)07:59
*** jpich has joined #openstack-infra07:59
openstackgerritNatal Ngétal proposed openstack/diskimage-builder master: [Configuration] Switch to stestr.  https://review.opendev.org/62941408:12
*** ykarel|lunch is now known as ykarel08:14
*** pkopec has joined #openstack-infra08:15
*** Emine has joined #openstack-infra08:20
*** tkajinam has quit IRC08:35
*** e0ne has joined #openstack-infra08:40
*** ykarel is now known as ykarel|meeting08:43
*** ricolin_ has joined #openstack-infra08:45
*** ricolin_ has quit IRC08:47
*** ricolin_ has joined #openstack-infra08:47
*** ricolin has quit IRC08:48
*** rfarr has quit IRC08:55
*** udesale has joined #openstack-infra08:55
*** rfarr has joined #openstack-infra08:55
*** apetrich has joined #openstack-infra08:59
*** ykarel has joined #openstack-infra09:05
*** ykarel|meeting has quit IRC09:05
*** ricolin_ is now known as ricolin09:07
*** ykarel_ has joined #openstack-infra09:09
*** ykarel_ is now known as ykarel|meeting09:10
*** ykarel has quit IRC09:10
*** ykarel|meeting is now known as ykarel09:11
*** ricolin has quit IRC09:20
*** bhavikdbavishi has quit IRC09:27
*** yamamoto has quit IRC09:31
*** Lucas_Gray has joined #openstack-infra09:35
*** ykarel has quit IRC09:40
*** yamamoto has joined #openstack-infra09:40
*** adrianreza has quit IRC09:48
*** whoami-rajat has joined #openstack-infra09:52
*** ociuhandu has joined #openstack-infra09:55
*** roman_g has joined #openstack-infra09:57
*** ociuhandu has quit IRC10:00
*** ociuhandu has joined #openstack-infra10:01
*** kjackal has quit IRC10:08
*** Wryhder has joined #openstack-infra10:15
*** Lucas_Gray has quit IRC10:16
*** Wryhder is now known as Lucas_Gray10:16
*** jaosorior has joined #openstack-infra10:17
*** yamamoto has quit IRC10:21
*** bhavikdbavishi has joined #openstack-infra10:36
*** bhavikdbavishi has quit IRC10:38
*** kjackal has joined #openstack-infra10:44
*** lucasagomes has quit IRC10:49
*** yamamoto has joined #openstack-infra10:55
*** bhavikdbavishi has joined #openstack-infra10:57
*** yamamoto has quit IRC11:00
*** lucasagomes has joined #openstack-infra11:01
*** Lucas_Gray has quit IRC11:02
*** Lucas_Gray has joined #openstack-infra11:02
*** yamamoto has joined #openstack-infra11:05
*** yamamoto has quit IRC11:11
openstackgerritinspurericzhang proposed openstack/diskimage-builder master: Mailing lists change openstack-dev to openstack-discuss  https://review.opendev.org/66843511:17
openstackgerritMarios Andreou proposed openstack/diskimage-builder master: WIP: Adds release->rhel8 for ironic-agent pkg-map  https://review.opendev.org/66843611:18
*** rascasoft has quit IRC11:30
*** rascasoft has joined #openstack-infra11:33
*** pcaruana has quit IRC11:36
*** bhavikdbavishi has quit IRC11:37
openstackgerritinspurericzhang proposed openstack/reviewstats master: Mailing lists change openstack-dev to openstack-discuss  https://review.opendev.org/66844611:38
*** bhavikdbavishi has joined #openstack-infra11:44
*** yamamoto has joined #openstack-infra11:47
*** ykarel has joined #openstack-infra11:49
*** ykarel is now known as ykarel|meeting11:49
*** rh-jelabarre has joined #openstack-infra11:50
openstackgerritSimon Westphahl proposed zuul/nodepool master: Allow proceeding with requests on quota exceeded  https://review.opendev.org/66737111:51
*** jaosorior has quit IRC11:56
*** jaosorior has joined #openstack-infra11:56
*** jaosorior has quit IRC11:57
*** jaosorior has joined #openstack-infra11:57
*** dchen has quit IRC11:59
mordredmorning AJaeger - getting back up to speed, anything you want me to look at while I'm waking up?12:03
sshnaidm|ruckI have an error with docker proxy I suppose: HTTPError: 401 Client Error: Unauthorized for url: http://mirror.bhs1.ovh.openstack.org:8082/v2/tripleomaster/centos-binary-mistral-api/blobs/sha256:469cc40974973b41dc9d82f4cfe406a6c630d5952f86f819cd853d3785d4fc5f12:14
sshnaidm|ruckcould it be proxy failure or something else..?12:15
mordredsshnaidm|ruck: yeah - I'm still coffeeing, so I might be wrong, but what that usually means is that something has caused docker to talk to dockerhub instead of the proxy (the proxy hub takes credentials, and docker thinks the proxy is dockerhub, so if it falls back to _actual_ dockerhub it uses the credentials to do so which are not real credentials)12:18
*** rosmaita has joined #openstack-infra12:19
openstackgerritMarios Andreou proposed openstack/diskimage-builder master: WIP: Adds release->rhel8 for ironic-agent pkg-map  https://review.opendev.org/66843612:21
*** Lucas_Gray has quit IRC12:23
*** Lucas_Gray has joined #openstack-infra12:25
*** rlandy has joined #openstack-infra12:28
*** electrofelix has joined #openstack-infra12:30
*** piotrowskim has joined #openstack-infra12:31
*** tdasilva has joined #openstack-infra12:36
*** eharney has quit IRC12:38
*** Wryhder has joined #openstack-infra12:41
*** Lucas_Gray has quit IRC12:43
*** Wryhder is now known as Lucas_Gray12:43
*** ykarel|meeting is now known as ykarel12:50
*** rfarr_ has joined #openstack-infra12:53
sshnaidm|ruckis there a short way to retrieve all gate job names for a specific project? maybe via zuul api12:54
*** ekultails has joined #openstack-infra12:55
*** aaronsheffield has joined #openstack-infra12:55
*** rfarr has quit IRC12:56
*** yamamoto has quit IRC12:57
*** jaosorior has quit IRC13:03
*** jaosorior has joined #openstack-infra13:03
*** auristor has quit IRC13:05
*** Lucas_Gray has quit IRC13:06
*** Lucas_Gray has joined #openstack-infra13:07
*** yamamoto has joined #openstack-infra13:10
*** yamamoto has quit IRC13:14
*** eharney has joined #openstack-infra13:15
*** sthussey has joined #openstack-infra13:19
*** mriedem has joined #openstack-infra13:27
*** lseki has joined #openstack-infra13:29
*** iurygregory has quit IRC13:29
AJaegermorning mordred , the stack at https://review.opendev.org/#/q/status:open+++topic:forge-publish is something I'd like to see in ;)13:29
AJaegerconfig-core, please review also https://review.opendev.org/#/c/665012/ and https://review.opendev.org/66711913:31
mordredAJaeger: that whole stack looks nice13:34
AJaegerthanks, mordred - tobias-urdin did the work, I just split up ;)13:34
*** kjackal has quit IRC13:34
mordredAJaeger, tobias-urdin: well, it was very easy to follow, so thanks!13:35
AJaegermordred, corvus , I'd also like a  review for two changes to switch to opendev, see https://review.opendev.org/667820 and https://review.opendev.org/667813 , please?13:38
*** bhavikdbavishi has quit IRC13:39
tobias-urdinthanks for reviews and to AJaeger for completing it :)13:40
AJaegertobias-urdin, mordred , due to the depends-on, we need to merge them one-by-one - I'll add my own +A to move them forward13:42
mordredAJaeger: cool13:42
openstackgerritBogdan Dobrelya (bogdando) proposed openstack/openstack-zuul-jobs master: Add binary dependencies for building TeX docs  https://review.opendev.org/66846413:42
AJaegerthanks, mordred.13:43
*** kjackal has joined #openstack-infra13:44
openstackgerritMerged openstack/openstack-zuul-jobs master: Add openstack-tox-functional-py37 job  https://review.opendev.org/66711913:44
AJaegermordred: you might also be interested in fungi's changes for ansible-lint, see https://review.opendev.org/#/q/status:open+project:zuul/zuul-jobs++topic:ansible-lint . Feel free to -1 those you disagree with13:44
openstackgerritSean McGinnis proposed openstack/openstack-zuul-jobs master: Set python_version: 3.7 for openstack-tox-functional-py37  https://review.opendev.org/66846613:45
openstackgerritMerged openstack/project-config master: Remove nodepool builder config for control plane images  https://review.opendev.org/66501213:47
*** yamamoto has joined #openstack-infra13:50
sshnaidm|ruckI'd like to add configure-mirrors role to tox-molecule job to use docker mirros13:52
sshnaidm|ruckshould I create a different parent for this job that uses this role or what is better to do?13:53
sshnaidm|ruckzbr, clarkb ^^13:53
openstackgerritMerged zuul/zuul-jobs master: Prepare: Use PDK to build puppet module  https://review.opendev.org/66759013:53
AJaegersshnaidm|ruck: Isn't configure-mirror used by all tox jobs?13:54
*** jcoufal has joined #openstack-infra13:54
zbrsshnaidm|ruck: enabling support for mirrors is likely to break it? if not altering the tox-molecule job may be best approach.13:54
sshnaidm|ruckAJaeger, how can I check it? I'm afraid I don't use mirror in molecule job13:54
mordredAJaeger: I have left votes and comments13:55
zbrbtw, I have a more generic question: why not using a generic http-proxy for caching? molecule can make use of http(s)_proxy by default.13:56
AJaegersshnaidm|ruck: check the log file ;)13:56
AJaegersshnaidm|ruck: if you cannot find it, please point me to a review13:57
AJaegermordred: great13:57
*** yamamoto has quit IRC13:57
sshnaidm|ruckAJaeger, for example this job: http://logs.openstack.org/66/668166/3/check/tripleo-ansible-centos-7-molecule-tripleo-bootstrap/5672866/  how can I figure out that I used docker proxy (or not)14:00
sshnaidm|ruckAJaeger, I see configure-mirrors is running, but how to be sure it's active for docker14:02
AJaegersshnaidm|ruck: can't help with the "usage" part. Is there a way to query docker for a proxy?14:04
*** dpawlik has quit IRC14:05
zbrsshnaidm|ruck: is not, AFAIK docker needs individual config to use http proxy, it does *not* make use of http_proxy env vars (99.9% about that)14:05
zbrneeds to be added manually in its json config14:05
zbrbut this feature should be part of the role that configures docker, not needed to be separated from it IMHO14:05
openstackgerritMerged openstack/openstack-zuul-jobs master: Install PDK dependencies for puppet building  https://review.opendev.org/66758914:06
zbri have a ? reg use of requirement-check which apparently is called only when some req*.txt files are touched. and apparently you discover that they are broken only when you have to modify one of the files. faced this recently. any good solution for avoiding that? enabling requirements-check on any change seems like a waste of resources to me...14:06
*** pcaruana has joined #openstack-infra14:06
openstackgerritMerged openstack/project-config master: Install PDK dependencies for puppet building  https://review.opendev.org/66759514:07
AJaegercorvus: thanks for reviews, I'll send out an announcement about https://review.opendev.org/66781314:07
openstackgerritMerged zuul/zuul-website master: Change git to opendev.org URLs  https://review.opendev.org/66782014:08
clarkbzbr we dont usr a generic http proxy because we cant filter source addresses on the proxy and would have to be a wideopen proxy on the internet which we dont want14:09
clarkbinstead we reverse proxy things that likely wont have problems being proxied14:10
*** pgaxatte has quit IRC14:10
clarkb(we cant filter source addrs because we dont know what the valid ip range for each cloud region is)14:10
zbrclarkb: clouds made that much harder. maybe enabling auth could be a solution? somehow I find the need to re-configure and maintain more and more mirrors as an activity that does not scale very well.14:12
*** pgaxatte has joined #openstack-infra14:12
corvusthere is no auth that we could keep secret14:12
AJaegercorvus: email send to zuul-announce...14:14
corvus(conceivably, we could perhaps put a proxy on a private neutron network that we attach every instance to, but i'm not sure whether we would be able to do that in every cloud region that we use)14:15
clarkbya I think rax and ovh wouldnt be able to do that14:19
clarkb(but may be wrong)14:19
*** kaiokmo has quit IRC14:19
*** iurygregory has joined #openstack-infra14:19
*** jamesmcarthur has joined #openstack-infra14:20
*** hemna has quit IRC14:22
AJaegerclarkb: are you around and want to watch ianw's new opendev mirrors, see https://review.opendev.org/663854 for change, please14:22
clarkbAJaeger: thanks, I do intend on catching up on that today as I'll need to deploy one as soon aa donnyd reports fortnebula is ready again14:26
corvusAJaeger: approved14:27
corvus(the email)14:27
AJaegerstrange, http://zuul.openstack.org/stream/a52ccb5ff6124ea0a01fe62995330707?logfile=console.log gives me since several minutes only "End of stream", repeated every minute? Are those repeated node start failures for change https://review.opendev.org/627534 ?14:27
AJaegerthanks, corvus14:28
openstackgerritMerged zuul/zuul-jobs master: Use PDK to build puppet module  https://review.opendev.org/62753414:28
AJaegerah, now it's finished....14:28
AJaegerstrange - log file exists, just streaming problem?14:28
fungiAJaeger: if it's a job which reboots the node or causes the console stream process to die for other reasons, that's a possible explanation14:29
openstackgerritMonty Taylor proposed zuul/zuul master: Spec: Add a Kubernetes Operator for Zuul  https://review.opendev.org/65918014:29
AJaegerfungi: I got log files in the end...14:30
openstackgerritMonty Taylor proposed zuul/zuul master: Spec: Add a Kubernetes Operator for Zuul  https://review.opendev.org/65918014:30
fungiwell, logs still get collected14:30
fungijust the console stream won't produce any output and you'll see "end of stream" there instead while the job runs14:31
fungiinfra-root: reminder, i'm out all week and not around the computer much14:31
mordredfungi: have fun! I enjoyed doing that last week14:31
AJaegerfungi: have fun! thanks for explanation14:31
donnydclarkb I am hopeful it will be up in the next hour or so14:33
donnydAlso can I ask you to switch back from raw to normal images14:33
donnydI got the storage servers to run 2x faster, just needed to use a different backend... Hence the reason for rebuild...14:34
*** udesale has quit IRC14:34
*** udesale has joined #openstack-infra14:34
*** dklyle has joined #openstack-infra14:37
clarkbdonnyd: sure, qcow2 images?14:37
clarkbI've got some meetings today and the kids have a dentist visit so I'll be juggling that too, but will do my best to get things back up and into a testable state14:38
clarkbfungi: enjoy the vacation time14:38
openstackgerritMerged openstack/openstack-zuul-jobs master: Remove bindep role for Puppet build job  https://review.opendev.org/62755414:40
openstackgerritAndreas Jaeger proposed zuul/zuul-jobs master: Test install-pdk-dependencies role  https://review.opendev.org/66833714:40
openstackgerritMerged openstack/project-config master: Remove bindep role for Puppet build job  https://review.opendev.org/62756614:42
openstackgerritAndreas Jaeger proposed openstack/project-config master: Add testing job to zuul-jobs  https://review.opendev.org/66848314:46
AJaegercorvus: here's the job addition ^14:46
*** ginopc has quit IRC14:46
AJaegerthanks for reviews!14:47
*** rajinir has joined #openstack-infra14:55
*** pcaruana has quit IRC14:56
*** cmorpheus is now known as cmurphy14:56
openstackgerritMonty Taylor proposed zuul/zuul master: Spec: Add a Kubernetes Operator for Zuul  https://review.opendev.org/65918014:58
AJaegermordred: want to review https://review.opendev.org/668337 and https://review.opendev.org/668483 as well, please? A small job to test one of the two just added roles14:59
openstackgerritMonty Taylor proposed zuul/zuul master: Spec: Add a Kubernetes Operator for Zuul  https://review.opendev.org/65918015:01
donnydclarkb: yea, qcows will probably be better for internety things anyways15:01
*** ykarel is now known as ykarel|away15:02
donnydclarkb: no worries. Thank you for continuing to be helpful in getting this thing running well15:02
*** auristor has joined #openstack-infra15:05
*** josephrsandoval has joined #openstack-infra15:06
clarkbdonnyd: no problem and really thank you for helping us too. Just let me know when I should reupload images and run the nova unittests again15:08
*** diablo_rojo has joined #openstack-infra15:09
*** factor has joined #openstack-infra15:09
*** icarusfactor has quit IRC15:10
*** icarusfactor has joined #openstack-infra15:11
*** factor has quit IRC15:11
donnydclarkb: Is there a place I could download the image and then run a script or cloud-config?15:12
clarkbdonnyd: there is! one sec15:12
donnydNice!!!15:12
clarkbdonnyd: https://nb01.openstack.org/images/ and https://nb02.openstack.org/images/ have our imagse you can download them and confirm the hashes, then reupload and boot them. Note you need to set config drive to true and it will configure ssh keys for the root user15:13
clarkbI think https may not have a properly signed cert on those hosts. ianw ^ we should totally LE those now that lE is a thing we can do15:13
*** pgaxatte has quit IRC15:14
*** factor has joined #openstack-infra15:14
*** icarusfactor has quit IRC15:15
donnydAnd what nova unittests do you run? Just want to make sure I am running the metrics against the same thing you were doing last week15:17
clarkbdonnyd: last week I ran the full nova unittest suite with tox. The basic process is git clone nova, run bindep `/usr/bindep-env/bin/bindep -f /path/to/nova/bindep.txt test`, take that output and apt-get install the results, run nova/tools/test-setup.sh, then run tox -epy27 in nova dir15:18
*** ricolin has joined #openstack-infra15:20
*** ijw has joined #openstack-infra15:20
donnydOk I will give it a spin15:20
clarkbdonnyd: all of this was ont he ubuntu-bionic image which is where we run those tests with zuul15:21
*** icarusfactor has joined #openstack-infra15:21
*** factor has quit IRC15:24
*** lucasagomes has quit IRC15:31
*** ekultails has quit IRC15:31
*** ykarel|away has quit IRC15:35
*** goldyfruit has quit IRC15:39
*** jaosorior has quit IRC15:42
*** ijw_ has joined #openstack-infra15:47
*** ekultails has joined #openstack-infra15:48
*** ijw has quit IRC15:51
*** weshay is now known as weshay_PTO15:52
*** pcaruana has joined #openstack-infra15:53
*** tdasilva has quit IRC15:56
*** e0ne has quit IRC16:00
*** josephrsandoval has quit IRC16:01
*** whoami-rajat has quit IRC16:01
*** ramishra has quit IRC16:02
*** jpich has quit IRC16:10
*** ricolin has quit IRC16:13
*** ykarel|away has joined #openstack-infra16:13
*** ykarel|away has quit IRC16:18
*** Lucas_Gray has quit IRC16:19
*** mattw4 has joined #openstack-infra16:20
*** ijw_ has quit IRC16:21
*** goldyfruit has joined #openstack-infra16:21
*** eernst has joined #openstack-infra16:21
*** eernst has quit IRC16:23
*** eernst has joined #openstack-infra16:23
*** mattw4 has quit IRC16:25
*** ricolin has joined #openstack-infra16:25
*** eernst has quit IRC16:26
*** eernst has joined #openstack-infra16:29
*** rlandy is now known as rlandy|brb16:30
*** chandankumar is now known as raukadah16:34
*** rpittau is now known as rpittau|afk16:36
*** Emine has quit IRC16:39
*** mattw4 has joined #openstack-infra16:43
*** ardaguclu has joined #openstack-infra16:46
*** witek has quit IRC16:46
zbrcfp.openstack.org persistently gives "There was a problem with our server, please contact admin."16:47
zbrafter which it redirects me to a callback page that appears empty16:48
*** udesale has quit IRC16:48
openstackgerritKevin Carter (cloudnull) proposed zuul/zuul-jobs master: Fix docker install on fedora  https://review.opendev.org/66852716:52
*** ricolin has quit IRC16:53
*** gyee has joined #openstack-infra17:01
*** ijw has joined #openstack-infra17:06
clarkbzbr: you can file bugs for that at https://bugs.launchpad.net/openstack-org I'll also let them know17:09
*** mattw4 has quit IRC17:10
zbrclarkb: thanks for the hint. is there a specific channel for the subject? For example I was wondering when is someone added to /community/speakers -- marking yourself available does not list you there.17:11
clarkbzbr: it is run by foundation staff and not the infra team. I don't think they have achannel but are typically responsive to email and bug reports17:12
zbrthanks, i will use email in this case.17:12
jamesmcarthurzbr: there is no specific channel for speakers, but you can send bug requests/reports to speakersupport@openstack.org17:12
*** ociuhandu has quit IRC17:13
*** mattw4 has joined #openstack-infra17:13
zbri will send may questions there17:13
jamesmcarthurzbr: regarding the call back error, we had an issue that has subsequently been fixed.17:13
jamesmcarthurIt's possible you have a cached version. Could you try clearing your cache?17:13
*** rh-jelabarre has quit IRC17:13
clarkbdonnyd: just a note that I'm not able to hit keystone currently17:13
clarkb(I'm finally back to computer and was goign to upload a qcow2, let me know if I can help)17:14
*** rh-jelabarre has joined #openstack-infra17:15
donnydyea... still going through gyrations over here17:15
donnydLOL17:15
clarkbdonnyd: no worries. Ping me when you think it is ready for me to have a go at it17:15
donnydhttps://bugs.launchpad.net/tripleo/+bug/183485717:15
openstackLaunchpad bug 1834857 in tripleo "Podman fails to start Glance API container mounted with NFS backend" [Undecided,New]17:15
*** ralonsoh has quit IRC17:16
donnydI am working around it now... Fingers crossed it will on be a couple more hours17:16
donnydI surely will17:16
donnydI was able to get the images pulled, so they will be uploaded stat once its back up17:16
clarkbdonnyd: note that those images are refreshed by our tooling daily so no need toupload them as "permanent" images17:16
*** smarcet has joined #openstack-infra17:17
clarkb(and I"m happy to upload them myself for testing)17:17
clarkbmordred: the other item on my todo list is cleaning up the control plane cloud images from nodepool17:18
clarkbmordred: I didn't want to start really digging into that until you were around17:19
mordredclarkb: cool- I'm about to lunch, but will be around to help once I'm no longer lunching17:19
clarkbmordred: https://review.opendev.org/#/c/665018/ and its depends on are the things to look at I think17:19
clarkbenjoy lunch!17:19
* mordred is gonna lunch all the lunches17:22
*** bobh has joined #openstack-infra17:24
*** ardaguclu has quit IRC17:27
openstackgerritMerged zuul/zuul-jobs master: Fix docker install on fedora  https://review.opendev.org/66852717:29
openstackgerritMerged openstack/project-config master: Switch RAX ORD mirrors to new opendev.org mirrors  https://review.opendev.org/66385417:29
clarkbAJaeger: ^ fyi. i'll keep an eye on it17:30
*** efried has quit IRC17:30
openstackgerritMerged openstack/project-config master: Switch ovh/GRA1 to opendev.org mirror  https://review.opendev.org/66836117:30
*** ardaguclu has joined #openstack-infra17:32
AJaegerthanks, clarkb17:33
*** rfarr__ has joined #openstack-infra17:33
*** ardaguclu has quit IRC17:34
*** ardaguclu has joined #openstack-infra17:34
AJaegerconfig-core, could review two changes to finish the puppet build changes, please? https://review.opendev.org/668337 and https://review.opendev.org/668483 as well, please? These add a small job for testing one of the roles.17:35
*** rfarr_ has quit IRC17:36
*** rlandy|brb is now known as rlandy17:36
clarkbAJaeger: first is approved. I'll approve the other once the first merges (due to the depends on)17:37
AJaegerthanks, clarkb17:37
*** electrofelix has quit IRC17:38
*** rfarr_ has joined #openstack-infra17:40
*** rfarr__ has quit IRC17:43
*** xek has quit IRC17:45
*** jamesmcarthur has quit IRC17:47
*** ociuhandu has joined #openstack-infra17:53
openstackgerritMerged zuul/zuul-jobs master: Test install-pdk-dependencies role  https://review.opendev.org/66833717:54
*** psachin has quit IRC17:55
*** harlowja has joined #openstack-infra17:58
*** ociuhandu has quit IRC18:02
prometheanfireis there a reason the base debian element in dib installs cloud-init, cloud-initramfs-growroot, etc?18:03
clarkbprometheanfire: I'm guessing because they are needed for useful cloud images18:05
clarkbnote the minimal element does not18:05
clarkbalso the non minimal elements tend to pull the upstream images and update them18:05
clarkbso could just be from that18:05
prometheanfireminimal element is also not systemd18:07
*** bobh has quit IRC18:07
prometheanfirehttps://github.com/openstack/diskimage-builder/blob/master/diskimage_builder/elements/debian-minimal/root.d/75-debian-minimal-baseinstall#L6918:08
prometheanfireor do docs need updating...18:08
clarkbprometheanfire: it is on our VMs18:08
clarkbI don't think there is any avoiding it on debian18:08
prometheanfireya, looks like docs were not updated :|18:08
prometheanfireNote that this element installs systemd-sysv as the init system18:09
prometheanfirehttps://github.com/openstack/diskimage-builder/tree/master/diskimage_builder/elements/debian-minimal18:09
prometheanfirefor that, meh18:09
*** igordc has joined #openstack-infra18:09
openstackgerritMerged zuul/zuul master: Add extra-config-paths tenant config option  https://review.opendev.org/66821418:09
clarkbthats still systemd and runs units right? I don't know what all the options are for debian but my undersatnding is modern debian is systemd required18:10
clarkbbut ya usually the non minimal elements take an upstream image and modify it (so you get all their opinions) and the minimal element attempts to be minimal18:10
prometheanfireyep, it's just docs being out of date18:11
openstackgerritMerged openstack/project-config master: Add testing job to zuul-jobs  https://review.opendev.org/66848318:14
*** igordc has quit IRC18:14
*** igordc has joined #openstack-infra18:15
*** igordc has quit IRC18:17
*** jamesmcarthur has joined #openstack-infra18:24
*** diablo_rojo has quit IRC18:30
*** weifan has joined #openstack-infra18:32
*** eernst has quit IRC18:34
*** diablo_rojo has joined #openstack-infra18:37
*** eernst has joined #openstack-infra18:37
*** kaiokmo has joined #openstack-infra18:38
openstackgerritGraham Hayes proposed zuul/nodepool master: Implement an Azure driver  https://review.opendev.org/55443218:45
*** weifan has quit IRC18:49
clarkbmordred: https://review.opendev.org/#/c/667986/ is another one you might make a quick glance at. In rebuilding gitea06 I ended up in a corrupt database table sitaution and I think that change will avoid that in the future by only backing up (then restoring) the gitea db18:50
clarkbmordred: if there is a reason to keep backing up the other dbs in mysql let me know (otherwise I'll likely approve that change)18:50
*** ijw has quit IRC18:56
*** ijw has joined #openstack-infra18:56
*** jtomasek has joined #openstack-infra18:56
openstackgerritJames E. Blair proposed zuul/zuul master: Reload config on extra config files changes  https://review.opendev.org/66854818:58
*** sshnaidm|ruck is now known as sshnaidm|afk18:59
openstackgerritMatthew Thode proposed openstack/diskimage-builder master: set default sources conf for buster as it now has a release  https://review.opendev.org/66854919:00
*** ardaguclu has left #openstack-infra19:08
*** ociuhandu has joined #openstack-infra19:12
*** ijw_ has joined #openstack-infra19:31
openstackgerritMatthew Thode proposed openstack/diskimage-builder master: install gnupg2 by default in debian-minimal  https://review.opendev.org/66855419:34
*** ijw has quit IRC19:35
*** bobh has joined #openstack-infra19:39
*** weifan has joined #openstack-infra19:49
*** yamamoto has joined #openstack-infra19:50
clarkbprometheanfire: dpkg doesn't already pull that in? huh19:53
prometheanfireclarkb: nope, and it's needed early enough that it needs to happen there, though it could be installed in the 02-gpg pre-install.d file19:55
*** yamamoto has quit IRC19:55
openstackgerritJames E. Blair proposed zuul/zuul master: Reload config on extra config files changes  https://review.opendev.org/66854820:00
*** hamzy has quit IRC20:02
*** dciabrin has quit IRC20:03
*** hamzy has joined #openstack-infra20:07
*** jtomasek has quit IRC20:08
*** rh-jelabarre has quit IRC20:14
*** rfarr has joined #openstack-infra20:14
*** eharney has quit IRC20:16
*** rfarr_ has quit IRC20:16
*** rfarr_ has joined #openstack-infra20:17
*** ociuhandu has quit IRC20:20
*** ociuhandu has joined #openstack-infra20:20
*** rfarr has quit IRC20:21
*** rh-jelabarre has joined #openstack-infra20:21
*** Lucas_Gray has joined #openstack-infra20:21
openstackgerritJames E. Blair proposed zuul/zuul master: Reload config on extra config files changes  https://review.opendev.org/66854820:22
mordredclarkb: reading now20:23
mordredclarkb: wow, they keep making mysqldump harder and harder to use properly20:24
clarkbmordred: fwiw I worked around it by manually editting the dump to not restore anything but the gitea db20:25
clarkbso I think this should be fine20:25
clarkbBut also you know things about that stuff so figured I would aks20:25
mordredclarkb: +2 from me - want me to just +a it or want to babysit?20:26
clarkbmordred: go ahead and +a and I'll check the results tomorrow morning20:26
clarkb(it backs up in the middle of the night iirc)20:26
mordreddone20:27
clarkbmordred: thanks@!20:33
clarkband then on the control plane image cleanup. Mostly I wanted us to watch that happen as merging stuff results in it deleting images20:33
clarkbwell watch that happen and/or double check there isn't anything we want from that20:34
*** ociuhandu has quit IRC20:34
clarkbmordred: also I discovered we want toadd at least ping6 when we get to better spot iwth that20:35
clarkbsince launch node attempts to ping6 things after boot20:35
mordredclarkb: oh - yeah. ping6 seems like an important thing then :)20:37
mordredclarkb: I don't think there's anything we want from the images that are there, so it should be fine to delete them20:37
mordredbut yeah - watching is good :)20:37
*** ijw_ has quit IRC20:39
*** ijw has joined #openstack-infra20:40
*** pcaruana has quit IRC20:41
*** pkopec has quit IRC20:41
*** pkopec has joined #openstack-infra20:43
clarkbok today has been weird for me so probably won't be approving any of that today, but if people can review tomorrow should be good20:44
*** tdasilva has joined #openstack-infra20:45
*** jamesmcarthur has quit IRC20:45
*** pkopec has quit IRC20:46
*** pkopec has joined #openstack-infra20:47
* clarkb tries to clean up some older changes that zuul is unhappy with20:47
*** Lucas_Gray has quit IRC20:50
donnydclarkb: finally I am getting somewhere20:52
openstackgerritMerged opendev/system-config master: Only backup the gitea database on gitea hosts  https://review.opendev.org/66798620:53
donnydOnly took 8 hours of deployments to figure out what the issue was20:53
clarkbhrm ansible query if told to errors='ignore' doesn't return an empty list if it would otherwise error? but skip: true does except skip:true is deprecated?20:54
clarkbso how are you supposed to use query(first_found) with loop and ignore errors?20:54
*** rfarr_ has quit IRC20:55
*** rfarr has joined #openstack-infra20:55
*** Lucas_Gray has joined #openstack-infra20:55
openstackgerritKendall Nelson proposed opendev/storyboard master: Link development.rst to contributing.rst  https://review.opendev.org/64596020:55
*** jcoufal has quit IRC20:56
clarkbthe docs say it should return a blank list hoewver testing says that isn't the case20:57
*** eharney has joined #openstack-infra20:57
mordredclarkb: we have a cantrip for that - it's complicated - let me find you an example20:57
clarkbhttp://logs.openstack.org/10/667710/2/check/system-config-run-base/619c1e8/job-output.txt.gz#_2019-06-27_00_17_25_538312 failures for https://review.opendev.org/#/c/667710/2/playbooks/roles/unbound/tasks/main.yaml20:58
mordredclarkb: hrm. I can't find the example of the thing you're talking about20:59
mordredclarkb: I remember beating my head against that a year ago or something like that though21:00
mordredcorvus, dmsimard: do you remember how do use loop, lookup(first_found) and to just do nothing if nothing is found? ^^21:00
clarkbthat contradicts https://github.com/ansible/ansible/blob/devel/lib/ansible/plugins/lookup/first_found.py#L47-L5621:01
clarkbmordred: fwiw the code changes I made were based on the examples21:01
clarkband I'm just updating an existing loop to use the non deprecated form of the lookup21:01
clarkbbecause our ansible run all log is full of deprecation warnings now and I find them annoying when trying to monitor ansible progress21:01
clarkbmaybe q != query?21:02
*** jamesmcarthur has joined #openstack-infra21:03
mordredclarkb: ah - nod - I understand now. and ... maybe it's different?21:04
*** bobh has quit IRC21:04
mordredclarkb: so maybe try just using q and see if it works?21:04
clarkbk I'll push that up and see if it works21:04
corvusmordred, clarkb: is it skip=true like in https://opendev.org/opendev/system-config/src/branch/master/roles/exim/tasks/main.yaml#L1421:05
*** ekultails has quit IRC21:05
*** mattw4 has quit IRC21:05
clarkbcorvus: that is the deprecated form that will be removed in 2.1221:05
clarkbso ya that works, its just going away in 2.12 apparently and we get tons of logging about it21:06
*** mattw4 has joined #openstack-infra21:06
corvuswhat?  that was the *new* thing that replaced "with_first_found"21:06
mordredcorvus: there's apparently a new new thing21:06
*** hwoarang has quit IRC21:06
clarkbcorvus: the change I linked has deprecation notice details in its commit message21:06
* corvus vocalizes thoughts in a medium not subject to permanent recording21:06
*** hwoarang has joined #openstack-infra21:09
openstackgerritClark Boylan proposed opendev/system-config master: Cleanup deprecated ansible syntax  https://review.opendev.org/66771021:09
*** slaweq has quit IRC21:11
*** tesseract has quit IRC21:12
*** weifan has quit IRC21:15
*** weifan has joined #openstack-infra21:17
*** weifan has quit IRC21:20
*** mattw4 has quit IRC21:25
*** mattw4 has joined #openstack-infra21:25
haleybredrobot: about a month ago i had started looking at moving the fedora-latest job to f29 from f28 and barbican was one of the consumers.  any chance you can look at my test patch?  https://review.opendev.org/#/c/662543/ - looks like there's some pki package changes to do, hoping you'd have a suggestion21:25
*** dciabrin has joined #openstack-infra21:26
*** mattw4 has quit IRC21:28
clarkbok more data on the ansible loop thing. It does still fail using q so I think q == query. On closer inspection the same include OS specific tasks works on bridge (bionic) but not a trusty test node. The only os specific tasks are for redhat so both of them should fail the lookup then return []21:29
clarkbpossibly a python version problem?21:29
clarkbwe do use python3 on bridge iirc21:29
*** weifan has joined #openstack-infra21:33
*** kjackal has quit IRC21:39
dmsimardmordred: apologies for the latency, on PTO21:42
dmsimardThere's an argument to include_vars or with_first_found to skip if nothing is found iirc21:43
dmsimardIf skip is being deprecated that's news to me21:44
clarkbdmsimard: it is :)21:44
clarkbdmsimard: [DEPRECATION WARNING]: Use errors="ignore" instead of skip. This feature will be removed in version 2.12.21:45
clarkbunfortunately the behavior doesn't seem to be consistent with skip21:45
dmsimardI suppose it's arguably similar but it's not the same yeah21:45
*** bobh has joined #openstack-infra21:46
clarkbthe examples for first_found document that this should work but it does not appear to. On further investigating it may work on different platforms (possibly a python3 vs python2 behavior difference)21:47
dmsimardWould "ignore_errors: yes" on the task itself work ?21:48
clarkbonly if you want to ignore errors from the tasks when they run I think21:48
clarkbwhich we dont21:48
dmsimardAh, right21:50
dmsimardI need to go back to "celebrate" Canada day (aka moving day in Quebec) but I'll be back tomorrow21:51
clarkbno worries, this isn't a rush. Was something I ran into (the warnings) when trinyg to follow logs and decided cleaning them up was worthwhile21:51
ianwclarkb: not totally caught up, did you see my note on the skip too?21:53
openstackgerritJames E. Blair proposed openstack/project-config master: Load extra Zuul config from zuul-jobs  https://review.opendev.org/66856321:54
ianwinfra-root: also; https://review.opendev.org/#/c/667782/ is i think ready for review.  it moves the rsync mirror updates to a opendev.org host managed by ansible.  the idea is to get everything on it, but reprepro and all the gpg key handling etc will need some more extensive hacking.  i figure it's a good start21:56
clarkbianw: oh I completely missed htat21:56
clarkbianw: so it is indeed a bug in ansible21:56
openstackgerritKendall Nelson proposed opendev/storyboard master: Link development.rst to contributing.rst  https://review.opendev.org/64596021:57
*** tdasilva_ has joined #openstack-infra21:57
*** weifan has quit IRC21:58
clarkbianw: any idea if ansible is backporting that fix? seems like an important one if they are going to deprecate that functionality with skip21:58
ianwclarkb: yeah, the skip bit is/was ... we could ask for a backport21:58
ianwjinx :)21:58
ianwpabelanger: ^ might have the contacts to get a backport in a (more) expedient fashion?21:59
ianwhttps://github.com/ansible/ansible/pull/57038 direct link21:59
*** tdasilva has quit IRC22:00
openstackgerritMerged zuul/zuul master: Reload config on extra config files changes  https://review.opendev.org/66854822:00
clarkbalso thanks for the update on the afs mirrors. Waiting on donnyd to get the cloud back into a happy state then I can deploy a mirror there with openafs 1.8.322:01
clarkbthough may do some quick nova based perf testing first22:01
donnydStill working22:01
clarkbdonnyd: no rush22:02
donnydThis isn22:02
donnydwell I would like to move on to actually using it22:02
donnydLOL22:02
*** mattw4 has joined #openstack-infra22:03
clarkbwell I don't want you to feel you have to rush on our behalf, but yes we too are excited about using it :)22:03
ianwclarkb: cool.  dhowells mentioned that cachefiles issues probably wouldn't get much time until maybe mid july.  that seems to be the major stability issue22:04
ianwin the mean time if we can help out with details to get the vos release / kafs seeing servers offline possible race issue sorted, that's one less thing too22:04
ianwso we're certainly not in "set and forget" mode with kafs, but at least we're pointed in the right direction22:05
*** weifan has joined #openstack-infra22:05
donnydI am not sure when NFS stopped working OOB.. but it surely is giving me quite a rough time today22:06
donnydLOL22:06
*** bobh has quit IRC22:06
corvusclarkb: wow you are not kidding about the deprecation warnings22:08
corvusthe logs are just useless22:08
clarkbcorvus: yup :/22:08
*** weifan has quit IRC22:09
corvusit's the same 6 line deprecation warning every 0.05 seconds.22:09
corvusi'm not sure we're ready for this yet -- but maybe in the near future we could set deprecation_warnings=False in production, but in zuul, have jobs report warnings back to zuul when deprecation warnings show up in runs...22:11
corvus(still, don't fail the jobs, just have them show up in the reports)22:12
*** mriedem has quit IRC22:13
corvusthe extra config zuul changes have landed, i'm going to restart zuul22:13
corvusand the repl server is in, so no more cherry-picking required22:14
clarkbcorvus: that is an interesting idea. I think as long as we expose those warnings that would be fine. Maybe a separate log file even22:14
*** rh-jelabarre has quit IRC22:15
corvusclarkb: yeah, i'm thinking something that ends up in the report message so we know to go looking for them -- but given that we need to identify them in order to know to raise the alarm, making a separate log file for them should be easy22:15
*** rfarr_ has joined #openstack-infra22:16
*** hamzy has quit IRC22:18
*** rfarr has quit IRC22:19
*** pkopec has quit IRC22:22
*** rh-jelabarre has joined #openstack-infra22:22
*** weifan has joined #openstack-infra22:24
*** jamesmcarthur has quit IRC22:27
*** jamesmcarthur has joined #openstack-infra22:27
*** rfarr_ has quit IRC22:27
*** rfarr_ has joined #openstack-infra22:28
*** jamesmcarthur_ has joined #openstack-infra22:28
*** rcernin has joined #openstack-infra22:30
corvus#status log restarted all of zuul on commit c5090244dc608a1ef232edded5cf92bf753dbb1222:31
openstackstatuscorvus: finished logging22:31
*** jamesmcarthur has quit IRC22:32
*** tdasilva_ has quit IRC22:40
*** tdasilva_ has joined #openstack-infra22:41
*** rfarr_ has quit IRC22:44
*** smarcet has quit IRC22:44
*** rfarr_ has joined #openstack-infra22:44
clarkbianw: re the mirror update node change: is the concern with updating the reprepro stuff that it changes on eg bionic or that it will just be more effort to convert from puppet to ansible?22:46
*** rfarr_ has quit IRC22:46
*** rfarr_ has joined #openstack-infra22:46
*** rfarr_ has quit IRC22:48
*** rfarr_ has joined #openstack-infra22:49
clarkbalso double checking I understand the testing there. It will deploy a mirror update node that will be set up to talk to the real afs volumes but the keytabs won't auth?22:50
clarkbso everything but the actual cron entry execution is exercised22:50
corvusyeah, my read is that its basically a test that the files got put on disk...22:51
clarkbya left another note that we might want to also test cron was updated22:51
clarkbin any case +2'd it and if my understanding is correct feel free to approve22:52
corvusi think it would be okay to put in a real keytab for an account which didn't have any special access, but i don't think that gets us much farther, we still won't be able to run the real rsync and release commands22:52
clarkbya22:52
corvuswe could parameterize the location, but then having a second copy of all the data to rsync sounds pretty heavyweight.  so all in all, i think what's there is about as much testing as is practical22:53
*** aaronsheffield has quit IRC22:53
*** tkajinam has joined #openstack-infra22:53
corvus(especially since this is a service where, when it breaks, everything just stays the same)22:54
*** rlandy has quit IRC22:56
*** hamzy has joined #openstack-infra22:57
clarkbfirefox has decided it needs an entire cpu 100% of the time. So I'm gonna reboot and pick up some much needed updates. Back in a bit22:57
openstackgerritKendall Nelson proposed opendev/storyboard master: Link development.rst to contributing.rst  https://review.opendev.org/64596022:58
*** eernst has quit IRC22:59
*** hwoarang has quit IRC23:00
*** hwoarang has joined #openstack-infra23:01
*** eernst has joined #openstack-infra23:02
*** Lucas_Gray has quit IRC23:02
*** igordc has joined #openstack-infra23:03
*** eernst has quit IRC23:04
*** jamesmcarthur_ has quit IRC23:05
*** igordc has quit IRC23:05
*** igordc has joined #openstack-infra23:06
ianwclarkb: yeah, no so much a concern as bigger job converting the reprepro puppet23:06
ianwyeah, the test is really just making sure things are decoded properly at this point, certainly could do more23:07
*** tosky has quit IRC23:08
*** tdasilva has joined #openstack-infra23:10
*** tdasilva_ has quit IRC23:11
*** tdasilva has quit IRC23:11
*** tdasilva has joined #openstack-infra23:12
*** hwoarang has quit IRC23:14
*** hwoarang has joined #openstack-infra23:17
*** rh-jelabarre has quit IRC23:18
*** smarcet has joined #openstack-infra23:18
*** andreaf has quit IRC23:18
ianwcorvus: i think for 100% safety that should be fixed?  but i think you have to be unlucky to get things that trigger an escape.23:20
*** tdasilva has quit IRC23:20
ianwi guess values in unicode code-points; maybe keytabs don't even have that23:21
*** weifan has quit IRC23:22
*** tdasilva has joined #openstack-infra23:22
*** tdasilva has quit IRC23:22
*** tdasilva has joined #openstack-infra23:22
ianwit seems like "ansible vault" is the answer ... not sure I want to ask the question though :)23:23
clarkbthat seems half an answer23:23
clarkbthere is such a thing as binary data that isn't secret23:23
clarkbmight even be a constant23:23
ianwvery true.  i thought the idea of a "data:" analog to "content:" for copy might have been good.  i find that many responses to ansible issues on github don't exactly foster discussion23:26
*** dchen has joined #openstack-infra23:32
*** andreaf has joined #openstack-infra23:34
*** rfarr_ has quit IRC23:38
*** rfarr_ has joined #openstack-infra23:39
*** sthussey has quit IRC23:39
*** rfarr_ has quit IRC23:50
*** rfarr_ has joined #openstack-infra23:50

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