Wednesday, 2016-03-23

« Tuesday, 2016-03-22 Index Thursday, 2016-03-24 »
*** suro-patz has joined #openstack-oslo00:00
*** salv-orl_ has joined #openstack-oslo00:02
*** dims_ has joined #openstack-oslo00:03
*** suro-patz has quit IRC00:04
*** dims has quit IRC00:04
*** salv-orlando has quit IRC00:05
*** salv-orl_ has quit IRC00:07
*** amrith is now known as _amrith_00:15
*** achanda has quit IRC00:18
*** achanda has joined #openstack-oslo00:21
*** _amrith_ is now known as amrith00:25
*** mriedem has joined #openstack-oslo00:34
*** achanda has quit IRC00:40
*** krotscheck has joined #openstack-oslo00:42
*** sputnik13 has quit IRC00:43
*** gcb has quit IRC00:56
*** gcb has joined #openstack-oslo01:07
*** achanda has joined #openstack-oslo01:11
*** browne has quit IRC01:14
*** EinstCrazy has joined #openstack-oslo01:21
*** mtanino has joined #openstack-oslo01:23
*** mriedem has quit IRC01:31
*** MVenesio has joined #openstack-oslo01:34
*** salv-orlando has joined #openstack-oslo01:35
*** zz_dimtruck is now known as dimtruck01:38
*** MVenesio has quit IRC01:39
*** dims_ has quit IRC01:40
*** ozamiatin has joined #openstack-oslo01:41
*** salv-orlando has quit IRC01:42
*** yamahata has quit IRC01:43
*** suro-patz has joined #openstack-oslo02:19
*** dimtruck is now known as zz_dimtruck02:31
*** amrith is now known as _amrith_02:32
*** gcb has quit IRC02:33
*** salv-orlando has joined #openstack-oslo02:41
*** gcb has joined #openstack-oslo02:48
*** salv-orlando has quit IRC02:51
*** suro-patz has quit IRC02:54
*** gcb has quit IRC02:55
*** gcb has joined #openstack-oslo03:13
*** vilobhmm11 has joined #openstack-oslo03:16
*** yamahata has joined #openstack-oslo03:26
*** links has joined #openstack-oslo03:28
*** MVenesio has joined #openstack-oslo03:35
*** MVenesio has quit IRC03:39
*** jeckersb is now known as jeckersb_gone03:40
*** gcb has quit IRC03:40
*** jecarey has joined #openstack-oslo04:01
*** achanda has quit IRC04:06
*** ozamiatin has quit IRC04:07
*** salv-orlando has joined #openstack-oslo04:10
*** salv-orlando has quit IRC04:14
*** yamamoto_ has joined #openstack-oslo04:30
*** achanda has joined #openstack-oslo04:53
*** MVenesio has joined #openstack-oslo05:36
*** MVenesio has quit IRC05:40
*** salv-orlando has joined #openstack-oslo05:42
*** suro-patz has joined #openstack-oslo05:43
*** salv-orlando has quit IRC05:49
openstackgerritMerged openstack/oslo-specs: policy: Support YAML  https://review.openstack.org/27972505:59
*** achanda_ has joined #openstack-oslo06:13
*** achanda has quit IRC06:14
*** nkrinner has joined #openstack-oslo06:17
*** ihrachys has joined #openstack-oslo06:22
*** ihrachys has quit IRC06:31
*** suro-patz has quit IRC06:32
openstackgerritDinesh Bhor proposed openstack/oslo.db: Raise DbMigrationError instead of KeyError  https://review.openstack.org/29620906:47
*** salv-orlando has joined #openstack-oslo06:58
*** e0ne has joined #openstack-oslo06:58
*** ndipanov has quit IRC07:08
openstackgerritMerged openstack/oslo.messaging: Always set all socket timeouts  https://review.openstack.org/29407407:13
*** salv-orlando has quit IRC07:15
*** davidlenwell has quit IRC07:26
*** achanda_ has quit IRC07:26
*** e0ne has quit IRC07:28
*** davidlenwell has joined #openstack-oslo07:34
*** achanda has joined #openstack-oslo07:35
*** MVenesio has joined #openstack-oslo07:37
*** tesseract has joined #openstack-oslo07:41
*** tesseract is now known as Guest6084207:41
*** MVenesio has quit IRC07:41
*** dtardivel has joined #openstack-oslo08:00
*** achanda has quit IRC08:06
*** achanda has joined #openstack-oslo08:08
*** achanda has quit IRC08:11
*** achanda has joined #openstack-oslo08:12
*** nkrinner has quit IRC08:12
*** achanda has quit IRC08:12
*** gcb has joined #openstack-oslo08:12
*** EinstCra_ has joined #openstack-oslo08:13
*** achanda has joined #openstack-oslo08:13
*** achanda has quit IRC08:13
*** EinstCrazy has quit IRC08:16
*** nkrinner has joined #openstack-oslo08:17
*** fhermeni has joined #openstack-oslo08:28
fhermenihi08:28
*** ihrachys has joined #openstack-oslo08:30
openstackgerritMerged openstack/pylockfile: Remove oslo.concurrency from replacement recommendation  https://review.openstack.org/29414708:31
*** xek has quit IRC08:32
*** xek has joined #openstack-oslo08:33
*** shardy has joined #openstack-oslo08:37
openstackgerritThomas Herve proposed openstack-dev/pbr: Fix wsgiref script use with oslo.config  https://review.openstack.org/29606108:42
*** permalac has joined #openstack-oslo08:43
*** itisha has quit IRC08:59
*** links has quit IRC09:01
*** e0ne has joined #openstack-oslo09:03
*** openstackgerrit has quit IRC09:03
*** openstackgerrit has joined #openstack-oslo09:03
*** yassine has joined #openstack-oslo09:04
*** yassine is now known as Guest9254209:04
*** Guest92542 is now known as yassou09:05
*** salv-orlando has joined #openstack-oslo09:06
*** gcb has quit IRC09:11
*** browne has joined #openstack-oslo09:11
*** lucas-dinner is now known as lucasagomes09:13
*** achanda has joined #openstack-oslo09:14
*** links has joined #openstack-oslo09:17
*** dougs has quit IRC09:18
*** dougs has joined #openstack-oslo09:19
*** achanda has quit IRC09:20
*** gcb has joined #openstack-oslo09:24
*** yamahata has quit IRC09:27
*** browne has quit IRC09:34
*** MVenesio has joined #openstack-oslo09:37
*** MVenesio has quit IRC09:42
*** cdent has joined #openstack-oslo09:53
*** vilobhmm11 has quit IRC09:56
*** sdague has joined #openstack-oslo09:58
*** dims has joined #openstack-oslo10:00
*** links has quit IRC10:05
openstackgerritMehdi Abaakouk (sileht) proposed openstack/tooz: file: make python2 payload readable from python3  https://review.openstack.org/29183610:07
*** ndipanov has joined #openstack-oslo10:09
*** EinstCra_ has quit IRC10:10
openstackgerritDina Belova proposed openstack/osprofiler: Clean thread local profiler object after usage  https://review.openstack.org/29451510:18
*** links has joined #openstack-oslo10:21
openstackgerritDavanum Srinivas (dims) proposed openstack/oslo.messaging: Update CheckForLoggingIssues hacking rule from keystone  https://review.openstack.org/29636510:29
*** cdent has quit IRC10:31
*** e0ne has quit IRC10:35
openstackgerritDavanum Srinivas (dims) proposed openstack/oslo.messaging: [py34] replace file with io.open  https://review.openstack.org/29636610:37
openstackgerritAlexis Lee proposed openstack/oslo.log: Reload log_config_append config on SIGHUP  https://review.openstack.org/29356110:37
openstackgerritAlexis Lee proposed openstack/oslo.log: Replace logging._lock  https://review.openstack.org/29592710:37
*** fhermeni has quit IRC10:42
*** e0ne has joined #openstack-oslo10:45
*** cdent has joined #openstack-oslo10:54
*** pkholkin__ has quit IRC10:59
*** dims_ has joined #openstack-oslo11:05
*** dims has quit IRC11:06
*** achanda has joined #openstack-oslo11:19
*** achanda has quit IRC11:24
*** ozamiatin has joined #openstack-oslo11:30
*** MVenesio has joined #openstack-oslo11:38
*** MVenesio has quit IRC11:43
openstackgerritAlexis Lee proposed openstack/oslo-specs: config: YAML config  https://review.openstack.org/27346811:50
dougsDoes the introduction of Rabbit heartbeats in Oslo (https://review.openstack.org/#/c/167308/1) mean that it's safe to use HAProxy to load balance a Rabbit cluster? Or are there still corner cases?11:55
patchbotdougs: patch 167308 - oslo.messaging (stable/kilo) - rabbit: heartbeat implementation (MERGED)11:55
*** john_bar has joined #openstack-oslo11:55
dims_dougs : most operators we polled last are NOT using HAProxy. i don't know if anyone has done extensive performance and load balancing tests using HAProxy11:58
dims_dougs : do you use ceilometer notifications?11:58
*** david-lyle_ has joined #openstack-oslo11:58
dougsdims_: thanks, no, we don't.11:59
dims_dougs : so you don't use "Highly Available Queues". right?12:00
*** david-lyle has quit IRC12:00
odyssey4medougs in OpenStack-Ansible we setup a list of rabbitmq targets which are all setup in HA Queues, no load balancer involved12:00
odyssey4methat configuration has been used at large scale in the OSIC cluster, several RAX Private Cloud customers and other OSA deployers12:01
dims_dougs : if you are not using notifications and not using HA queues, then you may try HAProxy at your own risk if you test it well :)12:01
odyssey4meback before Icehouse the RPC team tried using a load balancer for Rabbit, and it didn't work very well12:01
dims_odyssey4me : +112:01
*** salv-orl_ has joined #openstack-oslo12:02
dougsdims_: You mean you shouldn't use mirroring because we only have RPC going over Rabbit?12:02
dims_dougs : y12:02
*** MVenesio has joined #openstack-oslo12:02
dims_dougs : https://review.openstack.org/#/c/277948/ (change in fuel)12:03
patchbotdims_: patch 277948 - fuel-library - Disable HA for RPC queues by default (MERGED)12:03
dougsodyssey4me: thanks, that's good to know.12:03
dougsdims_: thanks!12:03
dims_dougs : of course, newer rabbitmq + erlang + kombu + pyamqp is a must :)12:04
*** salv-orlando has quit IRC12:04
odyssey4medims_ ++ we're using https://github.com/openstack/openstack-ansible/blob/liberty/playbooks/roles/rabbitmq_server/defaults/main.yml#L2912:05
dims_odyssey4me : nice12:05
odyssey4medims_ interesting to see that fuel has opted to disable ha - what does that actually means in terms of OpenStack/Rabbit config?12:05
dims_odyssey4me : we differentiate between RPC and notifications12:06
dims_dmitryme : anything to add here? ^^ please see discussion above12:07
odyssey4medims_ how does that differentiation come down in terms of a config example, if you have one?12:08
odyssey4medims_ perhaps what's documented here may provide some reference: http://docs.openstack.org/ha-guide/controller-ha-rabbitmq.html12:09
*** _amrith_ is now known as amrith12:09
dims_odyssey4me : http://codesearch.openstack.org/?q=enable_rpc_ha&i=nope&files=&repos=12:09
dims_odyssey4me : specifically set_rabbitmq_policy.sh12:10
odyssey4medims_ hmm, so is this assuming that the same vhost is used by all services, and thus the policy is set on specific queues to be ha?12:11
dims_odyssey4me : y. waiting for dmitryme to confirm12:11
*** pblaho has quit IRC12:11
dims_probably not here today12:12
dims_or in a meeting :)12:12
*** lucasagomes is now known as lucas-hungry12:12
odyssey4meinteresting - we have quite a different approach - each service has its own vhost, we found that with ha enabled all the way for everything, splitting each service into vhosts really helped at scale12:12
dims_odyssey4me : we've had problems with destructive testing at 200 nodes. (shoot controller, block ports using iptables etc) hence we ended up with above12:14
dims_odyssey4me : are you talking about performance speed right?12:14
*** jaypipes has joined #openstack-oslo12:15
odyssey4medims_ performance, yes, but also the ha cluster load balancing works better when splitting it up like that12:15
dougsso all OpenStack services can transparently handle loosing RPC messages in the event of a queue failure?12:15
odyssey4meas I understand it, using a single vhost means that all messages are actually only processed by one member of the cluster - the rest are just mirrors in case of failure12:16
odyssey4medims_ but yeah, perhaps we should share better in the enterprise/scale work groups so that we can all learn from each other :)12:17
dims_odyssey4me : some of the mirantis folks are active in #openstack-performance where do you all hang out?12:18
dims_dougs : some in flight requests may fail12:19
odyssey4medims_ I'll encourage some of the #openstack-ansible folks who are working on scale to join there too and contribute to the conversation.12:19
*** pblaho has joined #openstack-oslo12:20
dmitrymeodyssey4me: right, Fuel sets HA policy on Notifications queues only. Starting from Mitaka it is done by default. Here is the actual logic - https://github.com/openstack/fuel-library/blob/master/files/fuel-ha-utils/ocf/set_rabbitmq_policy.sh12:20
*** yamamoto_ has quit IRC12:20
dmitrymeas you see, we pick Ceilometer queues by regex "^(event|metering|notifications)\."12:20
odyssey4medmitryme so the reasoning here is that RPC messages can be lost in the event of a cluster breakdown of some sort?12:21
odyssey4mewhereas notification messages are essential12:21
dmitrymeright, and we want to reduce load on RabbitMQ as much as possible12:22
dims_yep12:22
odyssey4methat's pretty neat, thanks!12:22
dims_odyssey4me : we have a team member contributing directly to rabbitmq development too fyi :)12:22
odyssey4mewe've tested in OSIC up to 1000 nodes and haven't needed it - we have done some other tuning which helped though - let me get a link12:22
dims_odyssey4me : that's great news12:23
odyssey4mehere you go: https://github.com/openstack/openstack-ansible-rabbitmq_server/commit/a90cd7e34395196768be476be7f72655ca9bccdc12:23
*** binarin has joined #openstack-oslo12:23
odyssey4meas I recall, with separated vhosts and those settings they could build rather large stacks using heat in under 10 secs with very low failure rates12:24
dims_binarin : do we have any tuning? odyssey4me pointed out to  https://github.com/openstack/openstack-ansible-rabbitmq_server/commit/a90cd7e34395196768be476be7f72655ca9bccdc12:24
odyssey4menote also that the rabbitmq servers were on shared controller hosts - not on dedicated hardware12:24
dmitrymeodyssey4me: hmm, interesting, thank you12:24
dims_odyssey4me : right12:24
dims_us too for the moment12:25
dmitrymeodyssey4me: and all queus were mirrored?12:25
odyssey4mewe do use lxc machine containers to separate requirements, and make it quicker to recover from issue in operations (blow away the container, rebuild)12:25
dims_dmitryme : yep that's what odyssey4me mentioned12:25
odyssey4medmitryme yep12:26
dmitrymehmm, cool12:26
dims_dmitryme : binarin : we should test their tuning params! and explore separate vhosts as well12:26
dims_1000 nodes is pretty good12:26
dims_thanks for sharing odyssey4me12:27
odyssey4medims_ dmitryme a pretty blanket rule for mirroring: https://github.com/openstack/openstack-ansible-rabbitmq_server/blob/master/tasks/rabbitmq_post_install.yml#L25-L3312:27
thervejd__, https://bugs.launchpad.net/pbr/+bug/150175612:27
openstackLaunchpad bug 1501756 in PBR "Can't set port on wsgi server" [Undecided,In progress] - Assigned to Thomas Herve (therve)12:27
dmitrymeI didn’t thought that splitting into vhosts increases performance12:28
*** edmondsw has joined #openstack-oslo12:28
dmitrymeI mean, that is quite surprising12:29
dims_dmitryme : learn something new everyday :)12:29
dmitrymedims: right :-)12:29
odyssey4medmitryme it'd be nice to have an independent validation of that :)12:30
binarinUnless there are a lot of policies involved, I don't see how vhost count could affect performance.12:30
odyssey4mebinarin I could be completely wrong and it makes no difference to performance. All I know is that the guys who tested got far better results.12:31
dims_binarin : i think we should try the tuning first12:31
odyssey4meThe additional tuning parameters gave much, much better results and came a few months later.12:31
dims_ah i see12:32
dims_odyssey4me : at least it's easy to throw in quickly12:32
dmitrymeodyssey4me: we will definitely try that, though probably not right now. I will let you know how that went12:32
dims_separate vhosts is more work to enable12:32
odyssey4mewe still need to explore the use of galera multicast cluster comms to see if that helps when scaling out the galera cluster, but that's another matter12:32
odyssey4medmitryme cool :)12:33
odyssey4meI've joined the #openstack-performance channel and will lurk there. :)12:33
*** pradk is now known as pradk_12:33
binarinasync thread count was bumped in rabbit 3.5.4 to 6412:33
dims_odyssey4me : ack. we were talking about the ZMQ driver performance etc there recently12:33
dougsodyssey4me, dims_, dmitryme and others, thanks a lot, i've got some reading to do..12:36
odyssey4mesure thing dougs :) there is no one-size-fits-all solution, so the best solution is the one that meets your needs :)12:36
dims_dougs : thanks for starting the conversation :)12:37
dims_++ odyssey4me12:37
binarin(Unscientific) recommendation from rabbitmq team is async pool of 16*number_of_cores, so probably we should bump it further. Or even calculate this value automatically.12:37
dmitrymeodyssey4me: btw, we have a experienced another problem - the majority of queues are created on a single RabbitMQ node after full RabbitMQ cluster restart. That is because RabbitMQ cluster always starts with one node and later other nodes are joined to the first. While other RabbitMQ nodes join, the majority of OpenStack services connect to the first RabbitMQ and create queues here. Did you see such issue in your environemnt?12:37
odyssey4medmitryme yes, absolutely12:38
odyssey4medmitryme I don't know the details and whether that issue has been resolved in some way, but I remember being told about it.12:38
odyssey4meOnce our US folks are online (soon), I'll suggest we have a discussion about that in the #openstack-performance channel. Hopefully the right people will be available.12:39
dmitrymesounds good, I am interested in that topic. Right now we think how to handle it and so far we do not see an easy solution12:40
*** EinstCrazy has joined #openstack-oslo12:42
openstackgerritMerged openstack/oslo.db: Add new filter for DBDataError exception  https://review.openstack.org/29509612:42
jd__therve: ah thanks I didn't check the bug -_-12:51
*** jecarey_ has joined #openstack-oslo12:51
therveI thought so :p12:51
jd__therve: tried to set argv to [] ?12:51
thervejd__, When?12:52
jd__therve: before calling the app()12:52
thervejd__, Then I can't pass other arguments?12:52
jd__just use argparse and reset argv to [] when you parsed it12:52
jd__err no12:52
jd__it's a WSGI app there's no need for arg right?12:52
therveconfig-dir at least is useful12:52
thervedebug too12:52
jd__therve: then maybe using argparse parse_known_args()12:53
*** mtanino has quit IRC12:53
jd__and use the remainder as sys.argv12:53
jd__options, remainder = parser.parse_known_args(argv)12:54
therveOK I can try that12:54
* jd__ crosses his fingers12:54
*** jecarey has quit IRC12:54
*** kgiusti has joined #openstack-oslo12:55
*** jecarey_ has quit IRC12:56
thervejd__, OK so it only breaks --help12:56
jd__lol12:56
therveOtherwise seems ok12:56
therveI kinda prefer my solution then :)12:56
jd__how does it break help exactly? it's caught by argparse?12:57
therveYeah12:57
jd__therve: you can check Aodh or Gnocchi for projects that does not rely on cfg.CONF12:57
jd__s/does/do/12:57
jd__hm12:57
jd__therve: special case help?12:57
thervejd__, They don't use wsgi_script though12:58
jd__therve: *yet*12:58
jd__:-P12:58
therveThen use it and fix it at that point :)12:58
jd__I don't see how to fix it with your patch actually12:58
jd__that's why I'm trying to make you write a different one :)12:58
therveI see where you're coming from, it's not a nice solution by any stretch12:58
thervebut it's only for testing purpose12:59
therveAre you sure it will break with my patch?12:59
therveI have more faith that it handles local/global cfg than oslo/non oslo13:00
*** links has quit IRC13:01
jd__therve: I'm not sure no13:01
jd__but since you register your option in cfg.CONF and that is never used in Gnocchi/Aodh, I can't see how it'd work13:03
jd__cfg.CONF() is never called13:03
therveFair enough13:04
therveSo I can remove the help from port13:04
therveI guess that's reasonable enough, and it leaves the interesting options documented13:04
*** david-lyle_ is now known as david-lyle13:06
binarindims_: https://github.com/rabbitmq/rabbitmq-server/issues/151 - I'll try to implement that async pool increase in rabbit upstream13:07
dims_binarin : awesome thanks13:08
dims_binarin : we could consider a local fix for MOS/Fuel as well13:08
binarindims_: The second tuned option (about max number of erlang processes) doesn't have performance impact, it just limits max number of simultaneously existing connections + channels + queues.13:08
*** rlrossit has joined #openstack-oslo13:08
openstackgerritIlya Shakhat proposed openstack/oslo.messaging: Simulator: make parameter wait_after_msg float  https://review.openstack.org/29574713:09
thervejd__, ^^13:09
dims_binarin : ack13:09
openstackgerritIlya Shakhat proposed openstack/oslo.messaging: Simulator: collect error stats  https://review.openstack.org/29572213:09
*** amrith is now known as _amrith_13:10
*** binarin has quit IRC13:11
*** yamamoto has joined #openstack-oslo13:12
*** lucas-hungry is now known as lucasagomes13:14
*** mriedem has joined #openstack-oslo13:18
*** MVenesio has quit IRC13:21
*** achanda has joined #openstack-oslo13:22
*** jeckersb_gone is now known as jeckersb13:25
*** achanda has quit IRC13:27
*** regXboi has joined #openstack-oslo13:32
openstackgerritOleksii Zamiatin proposed openstack/oslo.messaging: [zmq] Reduce number of connections  https://review.openstack.org/28709413:40
*** jecarey has joined #openstack-oslo13:44
*** jecarey has quit IRC13:44
*** mtanino has joined #openstack-oslo13:44
*** jecarey has joined #openstack-oslo13:45
*** gordc has joined #openstack-oslo13:49
*** mriedem is now known as mriedem_meeting13:55
*** pradk has joined #openstack-oslo14:01
*** sigmavirus24_awa is now known as sigmavirus2414:01
*** _amrith_ is now known as amrith14:02
*** dims has joined #openstack-oslo14:04
*** dims_ has quit IRC14:04
*** achanda has joined #openstack-oslo14:05
*** achanda has quit IRC14:07
*** salv-orl_ has quit IRC14:07
*** openstack has joined #openstack-oslo14:22
*** andymaier has joined #openstack-oslo14:40
*** mriedem_meeting is now known as mriedem14:43
*** dukhlov has quit IRC14:44
*** david-lyle_ is now known as david-lyle14:47
*** zzzeek has quit IRC14:50
*** yamahata has joined #openstack-oslo14:58
*** zzzeek has joined #openstack-oslo15:02
*** EinstCrazy has quit IRC15:03
*** gordc has quit IRC15:03
*** openstackstatus has joined #openstack-oslo15:12
*** ChanServ sets mode: +v openstackstatus15:12
openstackgerritDina Belova proposed openstack/osprofiler: Clean thread local profiler object after usage  https://review.openstack.org/29451515:15
openstackgerritDina Belova proposed openstack/osprofiler: Add hepler to trace sessions  https://review.openstack.org/29451615:15
*** browne has joined #openstack-oslo15:22
harlowja_at_homejd__,  i put tooz on https://etherpad.openstack.org/p/newton-cross-project-sessions fyi15:23
harlowja_at_homei'll write more, but i think we need to continue that still15:23
*** yamahata has quit IRC15:38
jd__harlowja_at_home: ack15:53
harlowja_at_homeDinaBelova, boris-42 do u want to add a osprofiler one to https://etherpad.openstack.org/p/newton-cross-project-sessions ?15:56
harlowja_at_homeseems like it might be useful to have one?15:57
DinaBelovaharlowja_at_home we may, although there should be dedicated Performance Team working group session - although we may have both :)15:58
*** ozamiatin has quit IRC15:58
harlowja_at_homek15:58
harlowja_at_homemore sessions the better15:58
harlowja_at_homeha15:58
DinaBelova:)15:58
harlowja_at_homeespecially if these ones have a wider audience15:58
harlowja_at_homewhich they might15:58
*** yamahata has joined #openstack-oslo16:00
*** jaypipes has quit IRC16:04
*** links has joined #openstack-oslo16:06
*** jaypipes has joined #openstack-oslo16:07
*** Guest60842 has quit IRC16:21
*** andymaier has quit IRC16:22
harlowja_at_homedims_, does http://status.openstack.org//openstack-health/#/job/periodic-tempest-dsvm-oslo-latest-full-master load for u?16:24
harlowja_at_homeseems sorta dead for me, lol16:25
*** sigmavirus24 is now known as sigmavirus24_awa16:29
*** salv-orlando has joined #openstack-oslo16:31
*** gordc has joined #openstack-oslo16:32
*** david-lyle has quit IRC16:33
*** david-lyle has joined #openstack-oslo16:33
*** zz_dimtruck is now known as dimtruck16:37
*** browne has quit IRC16:38
*** salv-orlando has quit IRC16:38
*** salv-orlando has joined #openstack-oslo16:42
*** links has quit IRC16:49
dims_mtreinish : ^^16:49
*** yassou has quit IRC16:50
*** salv-orlando has quit IRC16:54
*** achanda has joined #openstack-oslo16:58
*** achanda has quit IRC17:03
*** dims_ has quit IRC17:05
*** dims has joined #openstack-oslo17:06
*** achanda has joined #openstack-oslo17:14
*** ihrachys has quit IRC17:16
openstackgerritDavanum Srinivas (dims) proposed openstack/oslo.messaging: Claim python3 compatability for Newton onwards  https://review.openstack.org/29662217:18
*** dukhlov has joined #openstack-oslo17:19
dimsozamiatin_ : sileht : haypo : dukhlov : some patches to get back python3 compat for oslo.messaging - https://review.openstack.org/#/q/project:openstack/oslo.messaging+is:open+owner:%22Davanum+Srinivas+(dims)+%253Cdims%2540mirantis.com%253E%2217:20
*** suro-patz has joined #openstack-oslo17:22
haypodims: reviewed17:24
dimsthanks haypo17:25
*** rcernin has quit IRC17:29
*** achanda has quit IRC17:29
*** browne has joined #openstack-oslo17:31
*** achanda has joined #openstack-oslo17:33
*** tedross has joined #openstack-oslo17:33
*** doffm has left #openstack-oslo17:34
*** achanda has quit IRC17:41
*** tedross has quit IRC17:42
*** dukhlov has quit IRC17:45
*** yamamoto_ has quit IRC17:47
*** salv-orlando has joined #openstack-oslo17:48
*** mriedem1 has joined #openstack-oslo17:51
*** achanda has joined #openstack-oslo17:51
*** mriedem has quit IRC17:52
*** yamamoto has joined #openstack-oslo17:52
*** cdent has quit IRC17:53
*** mriedem has joined #openstack-oslo18:00
*** mriedem1 has quit IRC18:00
*** salv-orl_ has joined #openstack-oslo18:01
*** salv-orlando has quit IRC18:04
*** sigmavirus24_awa is now known as sigmavirus2418:06
*** lucasagomes is now known as lucas-dinner18:08
*** achanda has quit IRC18:09
*** krotscheck has quit IRC18:14
*** krotscheck has joined #openstack-oslo18:15
*** ericksonsantos has quit IRC18:16
*** cdent has joined #openstack-oslo18:19
*** yamamoto has quit IRC18:19
*** yamamoto has joined #openstack-oslo18:24
*** tedross has joined #openstack-oslo18:26
*** ericksonsantos has joined #openstack-oslo18:26
*** ihrachys has joined #openstack-oslo18:26
*** krotscheck has quit IRC18:26
*** krotscheck has joined #openstack-oslo18:26
*** jaypipes has quit IRC18:31
*** yamamoto has quit IRC18:33
*** yamamoto has joined #openstack-oslo18:33
*** yamamoto has quit IRC18:33
*** yamamoto has joined #openstack-oslo18:34
*** yamamoto has quit IRC18:37
*** jaypipes has joined #openstack-oslo18:38
*** krotscheck has quit IRC18:38
*** krotscheck has joined #openstack-oslo18:39
*** permalac has quit IRC18:39
*** yamamoto has joined #openstack-oslo18:44
*** yamamoto has quit IRC18:44
*** yamamoto has joined #openstack-oslo18:44
*** yamamoto has quit IRC18:45
*** tedross has quit IRC18:46
rbradforI'm seeking some help/feedback/suggestions on my attempt to remove openstack/common/apiclient. Have two questions about exceptions, and base.Resource. Have WIP at https://review.openstack.org/#/c/296667/.18:47
patchbotrbradfor: patch 296667 - python-troveclient - [WIP] Replace Oslo Incubator apiclient usage18:47
*** openstack has joined #openstack-oslo19:08
*** openstack has joined #openstack-oslo19:21
*** sigmavirus24 has quit IRC19:21
*** zhiyan_ is now known as zhiyan19:23
*** openstackstatus has joined #openstack-oslo19:23
*** ChanServ sets mode: +v openstackstatus19:23
*** DuncanT_ is now known as DuncanT19:24
*** therve has joined #openstack-oslo19:24
*** rlrossit has joined #openstack-oslo19:26
*** sigmavirus24 has joined #openstack-oslo19:26
*** alaski_ is now known as alaski19:28
*** achanda has joined #openstack-oslo19:35
*** jecarey has joined #openstack-oslo19:41
*** sreshetnyak has quit IRC19:42
*** dims_ has quit IRC19:43
*** dims_ has joined #openstack-oslo19:43
*** dims has quit IRC19:45
*** jecarey has quit IRC19:45
*** yamamoto has joined #openstack-oslo19:45
*** sreshetnyak has joined #openstack-oslo19:47
*** dims_ has quit IRC19:53
*** yamamoto has quit IRC19:53
*** dims has joined #openstack-oslo19:56
*** jecarey has joined #openstack-oslo19:57
*** yamahata has quit IRC20:04
*** ihrachys_ has joined #openstack-oslo20:04
*** ihrachys has quit IRC20:05
*** shardy_ has joined #openstack-oslo20:07
*** shardy has quit IRC20:08
*** e0ne has joined #openstack-oslo20:17
*** achanda has quit IRC20:18
*** openstack has joined #openstack-oslo20:31
*** ozamiatin has joined #openstack-oslo20:32
*** e0ne has quit IRC20:37
*** sdague has quit IRC20:44
*** kgiusti has left #openstack-oslo20:50
openstackgerritayoung proposed openstack/oslo-specs: Policy merge tool  https://review.openstack.org/29678520:54
*** shardy_ has quit IRC20:55
*** jeckersb is now known as jeckersb_gone20:56
*** david-lyle has quit IRC20:59
*** ozamiatin has quit IRC21:00
*** achanda has joined #openstack-oslo21:01
*** david-lyle has joined #openstack-oslo21:05
*** jaypipes has quit IRC21:06
*** _amrith_ is now known as amrith21:06
*** dtardivel has quit IRC21:08
*** amrith is now known as _amrith_21:20
*** david-lyle has quit IRC21:21
*** edmondsw has quit IRC21:22
*** edmondsw has joined #openstack-oslo21:22
*** edmondsw has quit IRC21:22
*** david-lyle has joined #openstack-oslo21:23
*** regXboi has quit IRC21:31
*** _amrith_ is now known as amrith21:32
*** tedross has joined #openstack-oslo21:35
*** dukhlov has joined #openstack-oslo21:36
*** jeckersb_gone is now known as jeckersb21:40
*** browne has quit IRC21:41
*** amrith is now known as _amrith_21:42
*** achanda has quit IRC21:46
*** tedross has left #openstack-oslo21:48
*** rlrossit has quit IRC21:51
harlowja_at_homedhellmann, redhat??? :-P21:53
harlowja_at_homeuh ohes21:53
harlowja_at_homelol21:53
*** mriedem has quit IRC21:53
*** ihrachys_ has quit IRC21:56
*** browne has joined #openstack-oslo21:57
*** sigmavirus24 is now known as sigmavirus24_awa21:59
dhellmannharlowja_at_home : yep22:04
*** achanda has joined #openstack-oslo22:16
*** dimtruck is now known as zz_dimtruck22:23
*** pradk has quit IRC22:45
*** yamahata has joined #openstack-oslo22:54
*** browne has quit IRC22:56
*** mriedem has joined #openstack-oslo23:01
*** browne has joined #openstack-oslo23:04
*** alaski has quit IRC23:05
*** dukhlov has quit IRC23:08
*** cdent has quit IRC23:09
*** alaski has joined #openstack-oslo23:11
*** gordc has quit IRC23:14
gusRequest for cores to review privsep switch from json->msgpack: https://review.openstack.org/#/c/295083/23:17
patchbotgus: patch 295083 - oslo.privsep - Switch to msgpack for serialization23:17
gusI wrote this change *way* back and am rather embarrassed that I apparently never sent it for review.23:18
gus(it allows privsep to be bytes-vs-unicode clean, so is rather important)23:18
*** openstack has joined #openstack-oslo23:22
*** openstackstatus has joined #openstack-oslo23:23
*** ChanServ sets mode: +v openstackstatus23:23
*** vilobhmm11 has quit IRC23:35
*** vilobhmm11 has joined #openstack-oslo23:35
*** vilobhmm11 has quit IRC23:39
harlowja_at_homedhellmann, neats23:45
*** harlowja_at_home has quit IRC23:45
*** harlowja_at_home has joined #openstack-oslo23:46
harlowja_at_homegus, for that msgpack review u might want to look over https://review.openstack.org/#/c/291836/ just so u don't make the same mistake/bug23:46
patchbotharlowja_at_home: patch 291836 - tooz - file: make python2 payload readable from python323:46
harlowja_at_home^ also uses msgpack23:46
harlowja_at_home(from oslo.serialization)23:47
gusharlowja_at_home: interesting.23:52
gusharlowja_at_home: I'm using msgpack 'raw' for utf8-encoded strings and 'bin' for bytes.23:53
gusSo as far as I know, that will work across python versions, assuming the msgpack library continues to implement those two types consistently (as it appears to do).23:53
harlowja_at_homegus, ya, just something to check23:54
harlowja_at_homejust incase :)23:54
gus(also, cross python versions isn't as much of an issue for privsep, because I'm tightly coupled and always using the same python version on both ends of the Unix socket)23:54
harlowja_at_homek23:54
harlowja_at_homethe oslo.serialization code @ https://github.com/openstack/oslo.serialization/blob/master/oslo_serialization/msgpackutils.py#L32723:55
harlowja_at_homeuse_bin_type=True23:55
harlowja_at_homeanywaysssss23:55
gusI admit I don't understand why some of those strings are being converted to unicode in https://review.openstack.org/#/c/291836/23:56
patchbotgus: patch 291836 - tooz - file: make python2 payload readable from python323:56
harlowja_at_homesomehow i messed that up and sileht is fixing it23:56
harlowja_at_homeya, sileht hopefully knows some things23:56
guserk.  That change has things like "this was encoded iff the utf8 version is no longer the same as the original string"23:58
harlowja_at_homeya, its weird23:59
harlowja_at_homesooo just wanted to warn u ;)23:59
harlowja_at_homethe usage in tooz is restricted (only certain keys)23:59
« Tuesday, 2016-03-22 Index Thursday, 2016-03-24 »

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!