*** ykarel__ is now known as ykarel | 04:10 | |
opendevreview | Stephen Finucane proposed openstack/pbr master: Use importlib.metadata where possible https://review.opendev.org/c/openstack/pbr/+/954042 | 09:13 |
---|---|---|
opendevreview | Stephen Finucane proposed openstack/pbr master: Use packaging where possible https://review.opendev.org/c/openstack/pbr/+/954043 | 09:13 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Cleanup removal of 'build_sphinx' command doc https://review.opendev.org/c/openstack/pbr/+/954044 | 09:13 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Remove 'test' distutils command override https://review.opendev.org/c/openstack/pbr/+/954045 | 09:13 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Move fixtures to their own module https://review.opendev.org/c/openstack/pbr/+/954046 | 09:13 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Remove support for custom compilers https://review.opendev.org/c/openstack/pbr/+/954047 | 09:13 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Move distutils commands to _compat https://review.opendev.org/c/openstack/pbr/+/954048 | 09:13 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Move functional tests to their own directory https://review.opendev.org/c/openstack/pbr/+/954049 | 09:13 |
opendevreview | Stephen Finucane proposed openstack/pbr master: docs: Update usage guide and README https://review.opendev.org/c/openstack/pbr/+/949051 | 09:13 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Deprecate '[entry_points]' section https://review.opendev.org/c/openstack/pbr/+/949052 | 09:13 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Deprecate '[files]' section https://review.opendev.org/c/openstack/pbr/+/949053 | 09:13 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Deprecate various '[metadata]' options https://review.opendev.org/c/openstack/pbr/+/949054 | 09:13 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Deprecate various '[backwards_compat]' options https://review.opendev.org/c/openstack/pbr/+/949055 | 09:13 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Deprecate 'tests_require' option https://review.opendev.org/c/openstack/pbr/+/954050 | 09:13 |
opendevreview | Stephen Finucane proposed openstack/pbr master: tox: Stop calling 'setup.py' https://review.opendev.org/c/openstack/pbr/+/949188 | 09:13 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Use packaging where possible https://review.opendev.org/c/openstack/pbr/+/954043 | 11:00 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Cleanup removal of 'build_sphinx' command doc https://review.opendev.org/c/openstack/pbr/+/954044 | 11:00 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Remove 'test' distutils command override https://review.opendev.org/c/openstack/pbr/+/954045 | 11:00 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Move fixtures to their own module https://review.opendev.org/c/openstack/pbr/+/954046 | 11:00 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Remove support for custom compilers https://review.opendev.org/c/openstack/pbr/+/954047 | 11:00 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Move distutils commands to _compat https://review.opendev.org/c/openstack/pbr/+/954048 | 11:00 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Move functional tests to their own directory https://review.opendev.org/c/openstack/pbr/+/954049 | 11:00 |
opendevreview | Stephen Finucane proposed openstack/pbr master: docs: Update usage guide and README https://review.opendev.org/c/openstack/pbr/+/949051 | 11:00 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Deprecate '[entry_points]' section https://review.opendev.org/c/openstack/pbr/+/949052 | 11:00 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Deprecate '[files]' section https://review.opendev.org/c/openstack/pbr/+/949053 | 11:00 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Deprecate various '[metadata]' options https://review.opendev.org/c/openstack/pbr/+/949054 | 11:00 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Deprecate various '[backwards_compat]' options https://review.opendev.org/c/openstack/pbr/+/949055 | 11:00 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Deprecate 'tests_require' option https://review.opendev.org/c/openstack/pbr/+/954050 | 11:00 |
opendevreview | Stephen Finucane proposed openstack/pbr master: tox: Stop calling 'setup.py' https://review.opendev.org/c/openstack/pbr/+/949188 | 11:01 |
gibi | hi infra! Is there a tracker for the multi node job issue where nodepool allows cross provider worker placement preventing nova compute to talk to rabbit hosted in the other provider? I have a nova tracker for it https://bugs.launchpad.net/nova/+bug/2115980 | 12:56 |
gibi | there was some chatter yesterday on openstack-nova about it https://bugs.launchpad.net/nova/+bug/2115980 | 12:58 |
gibi | https://meetings.opendev.org/irclogs/%23openstack-nova/%23openstack-nova.2025-07-03.log.html#openstack-nova.2025-07-03.log.html#t2025-07-03T18:20:12 | 12:58 |
fungi | gibi: there's https://review.opendev.org/954064 which we should be auto-upgrading to within the next 24 hours | 12:59 |
gibi | fungi: thanks | 13:00 |
fungi | our zuul servers upgrade to the latest master branch state through weekly rolling restarts that begin around midnight utc on saturday mornings | 13:01 |
gibi | good to know | 13:02 |
frickler | iiuc this still doesn't guarantee locality 100%, though. maybe devstack needs to add a precheck to reject nodesets, possibly based on some latency cutoff? | 13:05 |
sean-k-mooney | frickler: we discussed this a little on the nova channel. routableity is more imporant then latency but thats aslo a factor as is open port access | 13:44 |
sean-k-mooney | frickler: to me this si just an infra bug as our exisaitng ci proivde always provide network connectivet and our jobs were built to depend on tha tas part of the contact | 13:45 |
sean-k-mooney | we can adapt the jobst compensate for the new capablites of zuul/node pools | 13:45 |
sean-k-mooney | but i would conder the current weaking of that guarentee to be a regression | 13:46 |
sean-k-mooney | but ya having a simpel role that is run as a preplaybook to reject the nodes and allow the job to be retired is a good first step | 13:47 |
sean-k-mooney | extendign it ot be abelto factor in latency or ports woudl be another good step. we generally rely on the default secuffity group in the provider openign up access to all port for vms in the same secuirty gorup | 13:48 |
opendevreview | Cyril Roelandt proposed openstack/project-config master: Add centos-10-stream nodepool image https://review.opendev.org/c/openstack/project-config/+/954136 | 14:15 |
fungi | and just to reiterate, the bug is fixed (unless you consider trying anyway instead of returning node_failure to be a buggy behavior as well), we're merely waiting for this weekend's automated upgrade to be running on it | 14:29 |
sean-k-mooney | fungi: i consider not requiring an explcity opt in at the node set to be a regression | 14:39 |
sean-k-mooney | but i think you state that its not enable by defualt | 14:40 |
sean-k-mooney | and we ahve elected to turn it on in our config somewhere | 14:40 |
sean-k-mooney | so tis not nessiarly a zull bug | 14:40 |
sean-k-mooney | but rather a bug in the current config given our existign jobs cannot tolerate having nodes form diffent providers in the same nodeset | 14:41 |
fungi | sean-k-mooney: afaik it is not currently a configurable option. we're in the process of (nearly done) replacing nodepool-launcher with zuul-launcher. the former could only boot a nodeset within a single provider, which made things like mixed-architecture multi-node jobs impossible. zl doesn't have that limitation, but tries to boot all nodes from a nodeset within the same | 14:48 |
fungi | provider unless that's deemed impossible for some reason (e.g. different architectures, nodeset is larger than the total quota for the largest provider, et cetera) | 14:48 |
sean-k-mooney | fungi: yep but the fact that it could only boot a nodeset within a singel provier was also a feature not a bug | 14:53 |
sean-k-mooney | i like the idea of being able to have multi architecture jobs or have jobs that have a contaienr registry rom openshfit that is used with vms form openstack to test something | 14:54 |
sean-k-mooney | but i really dont think that shoudl be allowe by default and there shoudl be some way of expressing a policy around that wehn defining nodesets | 14:55 |
sean-k-mooney | im not saying we shoudl break the lable absraction by the way | 14:55 |
sean-k-mooney | just a policy in the node set that can ssay if its ok for the nodes ot come form difent proviers or how to group them | 14:56 |
sean-k-mooney | today if you wanted to do a multi architure job you woudl havd to crfeate to provider jobs taht pasused and a 3 djob that depened on both that then used the nodes form the provider jobs to execute the multi arch job | 14:57 |
sean-k-mooney | i.e an arm provider job, x86 provider job and a test execuitr job that depended on both and uses the nodes form both to do whatever | 14:58 |
sean-k-mooney | its nice that you culd do that more simplely now | 14:58 |
fungi | from a broader zuul userbase (not just opendev/openstack) it was a hindrance, particularly because static nodes get a separate provider and the old model made jobs incorporating a static node and ephemeral nodes into a nodeset together impossible | 15:48 |
clarkb | it also meant you cannot mix kubernetes pods/namespaces and VMs | 17:27 |
clarkb | I think we should see if this is actually a problem after the fix | 17:28 |
clarkb | rather than jumping to assumptions and demanding changes beforehand | 17:28 |
clarkb | the current behavior was acknowledged as buggy, a fix was pushed, and will be deployed shortly | 17:28 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Move fixtures to their own module https://review.opendev.org/c/openstack/pbr/+/954046 | 17:59 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Remove support for custom compilers https://review.opendev.org/c/openstack/pbr/+/954047 | 17:59 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Move distutils commands to _compat https://review.opendev.org/c/openstack/pbr/+/954048 | 17:59 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Move functional tests to their own directory https://review.opendev.org/c/openstack/pbr/+/954049 | 17:59 |
opendevreview | Stephen Finucane proposed openstack/pbr master: docs: Update usage guide and README https://review.opendev.org/c/openstack/pbr/+/949051 | 17:59 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Deprecate '[entry_points]' section https://review.opendev.org/c/openstack/pbr/+/949052 | 17:59 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Deprecate '[files]' section https://review.opendev.org/c/openstack/pbr/+/949053 | 17:59 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Deprecate various '[metadata]' options https://review.opendev.org/c/openstack/pbr/+/949054 | 17:59 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Deprecate various '[backwards_compat]' options https://review.opendev.org/c/openstack/pbr/+/949055 | 17:59 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Deprecate 'tests_require' option https://review.opendev.org/c/openstack/pbr/+/954050 | 17:59 |
opendevreview | Stephen Finucane proposed openstack/pbr master: tox: Stop calling 'setup.py' https://review.opendev.org/c/openstack/pbr/+/949188 | 17:59 |
sean-k-mooney | clarkb: no one is demandieng changes but i am reporting a backwared incompatbale change in behavior that does violate one of the requirement we have for the networkign in our exisitng jobs. | 18:35 |
sean-k-mooney | clarkb: zuul is free ot evolve to meet the wirder needs of its users | 18:35 |
sean-k-mooney | but i would hope that not brekaign there users is also valued and that providing feedback when there are breakign changes is also | 18:36 |
sean-k-mooney | i have hit the exact limitation with mixing pods form openshfit and vms form opnestack and or static nodes | 18:37 |
sean-k-mooney | and i have also worked around tha tusing job depencies before i.e. where i spawned a pod to run a cli to provision bemetal host and then consomed those as static nodes | 18:38 |
sean-k-mooney | i.e. i spwaned a pod on openshift to run the beaker cli to provision static nodes becauses nodepool did not have a beaker driver | 18:38 |
sean-k-mooney | it takes a lot more understanding of how zuul and nodepool worked to be able ot leave multile providers in the past. as you had to intrally device thigs up into multipel jobs. with depenices. | 18:46 |
sean-k-mooney | nodeset.alternatives seam cool | 18:46 |
sean-k-mooney | i feel like thats an underused feature | 18:47 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Move functional tests to their own directory https://review.opendev.org/c/openstack/pbr/+/954049 | 20:35 |
opendevreview | Stephen Finucane proposed openstack/pbr master: docs: Update usage guide and README https://review.opendev.org/c/openstack/pbr/+/949051 | 20:35 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Deprecate '[entry_points]' section https://review.opendev.org/c/openstack/pbr/+/949052 | 20:35 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Deprecate '[files]' section https://review.opendev.org/c/openstack/pbr/+/949053 | 20:35 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Deprecate various '[metadata]' options https://review.opendev.org/c/openstack/pbr/+/949054 | 20:35 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Deprecate various '[backwards_compat]' options https://review.opendev.org/c/openstack/pbr/+/949055 | 20:35 |
opendevreview | Stephen Finucane proposed openstack/pbr master: Deprecate 'tests_require' option https://review.opendev.org/c/openstack/pbr/+/954050 | 20:35 |
opendevreview | Stephen Finucane proposed openstack/pbr master: tox: Stop calling 'setup.py' https://review.opendev.org/c/openstack/pbr/+/949188 | 20:35 |
Generated by irclog2html.py 4.0.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!