| opendevreview | Ghanshyam proposed openstack/tempest master: Add more server migration tests https://review.opendev.org/c/openstack/tempest/+/953847 | 01:06 |
|---|---|---|
| *** mhen_ is now known as mhen | 01:14 | |
| opendevreview | Milana Levy proposed openstack/tempest master: Adding a password regex test for PCI DSS https://review.opendev.org/c/openstack/tempest/+/954029 | 11:44 |
| opendevreview | Milana Levy proposed openstack/tempest master: Adding a password regex test for PCI DSS https://review.opendev.org/c/openstack/tempest/+/954029 | 11:54 |
| opendevreview | Milana Levy proposed openstack/tempest master: Adding a password regex test for PCI DSS https://review.opendev.org/c/openstack/tempest/+/954029 | 11:59 |
| opendevreview | Eduardo Olivares proposed openstack/devstack-plugin-ceph master: DNM - dummy test https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/954033 | 12:59 |
| opendevreview | Rajesh Tailor proposed openstack/tempest master: Add response schema for nova microversion v2.101 https://review.opendev.org/c/openstack/tempest/+/953331 | 16:37 |
| opendevreview | Rajesh Tailor proposed openstack/tempest master: Add response schema for nova microversion v2.101 https://review.opendev.org/c/openstack/tempest/+/953331 | 16:42 |
| opendevreview | Ghanshyam proposed openstack/tempest master: Add more server migration tests https://review.opendev.org/c/openstack/tempest/+/953847 | 17:35 |
| sean-k-mooney | gmaan: dumb question i did tempest run --serial -r ... | 17:58 |
| sean-k-mooney | i have 23 failures | 17:58 |
| sean-k-mooney | is there an easy way from me to find out which tests failed | 17:59 |
| clarkb | doesn't tempest run stestr under hood? you can probably do stestr failing --list or whatever the command is | 18:00 |
| sean-k-mooney | i think it does but it has a --subunit flag that i did not pass so i dont know if it wrote that db | 18:02 |
| sean-k-mooney | but its defintly worth a try | 18:02 |
| sean-k-mooney | .stestr/failing exist so that a good sign | 18:03 |
| gmaan | yeah, there you can see, tempest run does not have failing option in cmd | 18:04 |
| sean-k-mooney | clarkb ++ that works perfectly and i cna also get the subunit stream so i assume i could generate hte html report if i really wanted | 18:04 |
| clarkb | the --subunit flag should only affect output not database recording | 18:04 |
| sean-k-mooney | im prety sure some of these test are failing becasue my devstack cant actully run them | 18:08 |
| sean-k-mooney | like i know for the encyption test to pass i actully need to create speciic volume types and update my tempest config to sue them | 18:08 |
| sean-k-mooney | devstack get pretty close out of the box but its not perfect | 18:09 |
| sean-k-mooney | related question some of the test leak resouce which i normally clean up via horizon but i see there is also a tempest clenup command | 18:13 |
| sean-k-mooney | this is just a devstack but how safe is that in general | 18:13 |
| gmaan | yeah, ideally tests should not leak any resource that is bug in test but you can run cleanup | 18:13 |
| gmaan | it is safe, it only clean resources created by tempest. | 18:14 |
| sean-k-mooney | ok so does it just look for a tempest prefix or is it a littel smarter | 18:14 |
| gmaan | there are multiple way yon run it. save-state before tempest run and then run cleanup after tempest run so cleanup cmd will cleanup only resources diff from saved-state and what all resources it has after tempest run | 18:15 |
| gmaan | also prefix way you can run | 18:15 |
| gmaan | I prefer that one default prefix is tempest- | 18:15 |
| gmaan | you can run in --dry-run mode and see what all resource it will clean | 18:16 |
| sean-k-mooney | ack 99% of the stuff got cleaned up but one of the volume snapshto/backput tests is leakign the snapshot/backups so its not properly deleting them | 18:17 |
| sean-k-mooney | that might be related to the storage space i have alaibel or soemthing likethat | 18:17 |
| gmaan | k, feel free to file bug also test should be auto clean resources in any situation | 18:18 |
| sean-k-mooney | i think im hittign a cider issue not a tempest one | 18:18 |
| gmaan | ohk | 18:18 |
| sean-k-mooney | i mean it coudl be a bug in either or just but devstack env but to delete the voluem i have to force update a snaphsot then delete that with fornce and hten i can clean up the rest | 18:19 |
| sean-k-mooney | so i have not check to see if im just runnnig out os space on that backend or soemthing simple like aht | 18:19 |
| sean-k-mooney | normally i do not deploy with cinder backup so normally i dont run that test. | 18:21 |
| sean-k-mooney | ah Can not get Swift capabilities during backup driver initialization.: requests.exceptions.ConnectionError: HTTPConnectionPool(host='192.168.50.250', port=8080): Max retries exceeded with url: /info | 18:38 |
| sean-k-mooney | i do not have swift installedcx | 18:39 |
| sean-k-mooney | i guess that is what cidner backup uses by default | 18:39 |
| sean-k-mooney | so that explains why test_snapshot_backup is failing for me | 18:41 |
| sean-k-mooney | gmaan: when i run tempest form the tempet directory in devstack am i using /opt/stack/tempest/etc/tempest.conf ? | 18:44 |
| sean-k-mooney | i assume yes | 18:44 |
| sean-k-mooney | so i shoudl just be able to set service_available/swift=false | 18:45 |
| sean-k-mooney | well no CONF.volume_feature_enabled.backup=false woudl be better | 18:46 |
| sean-k-mooney | swift = False was set by devstack corectly | 18:47 |
| gmaan | sean-k-mooney: yeah, service enable/disable is controlled by devstack | 19:07 |
| gmaan | but if you just want test to skip then CONF.service_available.swift=false will work | 19:10 |
| *** profcorey is now known as Guest21311 | 19:30 | |
| opendevreview | Ghanshyam proposed openstack/tempest master: Add more server migration tests https://review.opendev.org/c/openstack/tempest/+/953847 | 21:22 |
| opendevreview | Ghanshyam proposed openstack/tempest master: Add multinode job for stable branches https://review.opendev.org/c/openstack/tempest/+/954070 | 21:30 |
| opendevreview | Ghanshyam proposed openstack/tempest master: Add multinode job for stable branches https://review.opendev.org/c/openstack/tempest/+/954070 | 21:33 |
| *** haleyb is now known as haleyb|out | 21:52 | |
| sean-k-mooney | gmaan: devstack did thei right thin and set swift as disabled however the default backend of cinder backend is swift so if you have that enabel wihtout swift then devstack does not configure a diffent backend so cinder backup does not work. the tempest test for backup dont assume the backend so they dont look at the service avaiable for swift which is correct | 23:47 |
| sean-k-mooney | so the bug is more that https://github.com/openstack/devstack/tree/master/lib/cinder_backups sercvice can be deploy without the relevnet backend. bug since it could be multi node devsack there is no way for devstack to know any better. in other word the porblem was with how i wrote my local.conf | 23:49 |
| sean-k-mooney | i alwasy turn of swift and cinder backup locall but this time i used a local.conf form a ci job and only turnd off swift | 23:50 |
Generated by irclog2html.py 4.0.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!