| *** tosky has quit IRC | 00:20 | |
| *** LinPeiWen15 has joined #openstack-manila | 00:49 | |
| *** ivan_lin has joined #openstack-manila | 00:49 | |
| *** LinPeiWen has quit IRC | 00:52 | |
| *** LinPeiWen15 has quit IRC | 01:13 | |
| *** LinPeiWen has joined #openstack-manila | 01:15 | |
| *** vishalmanchanda has joined #openstack-manila | 06:15 | |
| *** ivan_lin has quit IRC | 06:51 | |
| *** LinPeiWen has quit IRC | 07:15 | |
| *** LinPeiWen has joined #openstack-manila | 07:50 | |
| *** lpetrut has joined #openstack-manila | 08:44 | |
| *** tosky has joined #openstack-manila | 11:22 | |
| *** raildo has joined #openstack-manila | 11:48 | |
| *** raissa has joined #openstack-manila | 12:51 | |
| openstackgerrit | Lee Yarwood proposed openstack/devstack-plugin-ceph master: nova: Do not attempt to create vms pool when CEPH_REMOTE=true https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/756323 | 13:36 |
|---|---|---|
| openstackgerrit | Lee Yarwood proposed openstack/devstack-plugin-ceph master: WIP zuul: Introduce a multinode ceph job https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/711625 | 13:36 |
| *** raissa has quit IRC | 14:21 | |
| *** lpetrut has quit IRC | 15:03 | |
| openstackgerrit | Tom Barron proposed openstack/devstack-plugin-ceph master: Enable cephFS snapshots in manila https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/741275 | 19:11 |
| *** raissa has joined #openstack-manila | 20:05 | |
| *** raildo has quit IRC | 21:07 | |
| openstackgerrit | Tom Barron proposed openstack/devstack-plugin-ceph master: Bump NFS Ganesha version https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/756127 | 21:14 |
| openstackgerrit | Tom Barron proposed openstack/devstack-plugin-ceph master: Clean up CEPH_DATA_DIR when unstacking https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/765877 | 21:14 |
| openstackgerrit | Tom Barron proposed openstack/devstack-plugin-ceph master: Clean up CEPH_DATA_DIR when unstacking https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/765877 | 21:22 |
| openstackgerrit | Tom Barron proposed openstack/devstack-plugin-ceph master: Allow non-default Ceph cluster name https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/765659 | 21:22 |
| openstackgerrit | Tom Barron proposed openstack/devstack-plugin-ceph master: DNM -- Test alternate Ceph Cluster name https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/765878 | 21:22 |
| openstackgerrit | Goutham Pacha Ravi proposed openstack/manila stable/queens: [DNM] Testing tempest against this stable branch https://review.opendev.org/c/openstack/manila/+/765880 | 21:24 |
| *** raissa has quit IRC | 21:43 | |
| *** jmlowe has quit IRC | 21:45 | |
| *** jmlowe has joined #openstack-manila | 21:47 | |
| tbarron | vkmc: I think you mentioned that the Ceph NFS job failed test cases pertain to IPv6? If so, | 21:57 |
| tbarron | this may be a clue: https://zuul.opendev.org/t/openstack/build/a6fa338885ab41a08daad20a403c723b/log/controller/logs/devstacklog.txt#4967 | 21:57 |
| tbarron | vkmc: we use radvd to announce the return routes to self-service tenant networks with IPv6 | 21:58 |
| vkmc | tbarron, I saw the service is down | 21:58 |
| vkmc | but haven't inspected too much there | 21:59 |
| tbarron | vkmc: lemme look at the LVM jobs, cause they should have the same need | 21:59 |
| vkmc | tbarron, sounds good :) | 21:59 |
| tbarron | vkmc: also, ty for mentioning/suggesting that we may have the issue with ganesha exports and | 22:00 |
| tbarron | non-default cluster name with upstream devstack as well | 22:01 |
| vkmc | tbarron, yes... that's another thing we need to fix | 22:01 |
| tbarron | upstream devstack actually makes it hard to set things up like that, but I've hacked it | 22:01 |
| tbarron | in https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/765878 and | 22:02 |
| tbarron | https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/765659 | 22:03 |
| tbarron | vkmc: working on that I saw that you are right that the Ceph community is actually saying | 22:03 |
| tbarron | "don't name your cluster anything but 'ceph'" | 22:03 |
| tbarron | vkmc: I kinda scoffed at that when you said use of non-ceph name might be a doc issue, | 22:04 |
| tbarron | but apparently that is their stance ! | 22:04 |
| tbarron | That said, ceph-ansible still allows for use of a different name, and does this normally whe RGW is in the mix | 22:05 |
| tbarron | So we still need to figure out the ganesha issue. | 22:05 |
| vkmc | tbarron, yes... we need to change that, somehow... or at least properly document it | 22:05 |
| gouthamr | does the name of the cluster matter as much as the path to the conf file? | 22:06 |
| tbarron | Good news is that the problem reproduces in devstack where it should be easier to play with the exports format than in frozen container code. | 22:06 |
| gouthamr | nice ^ | 22:06 |
| vkmc | tbarron, not sure how many people tinkered with those ceph-ansible values | 22:06 |
| tbarron | gouthamr: dunno for sure, by convention (like ansible) the path to the conf file and the names of the cli9ent keys refelct the "cluster name" | 22:06 |
| vkmc | gouthamr, we need to test that | 22:07 |
| tbarron | The last may not have any real meaning beyond these except that it also shows up in the systemd units (at least in ubuntu) | 22:07 |
| tbarron | so you have to override those as well if you change it. | 22:08 |
| tbarron | by "the last" I mean the "cluster name" which the ceph people are trying to say is not useful any more, so they want to always let it be "ceph" and | 22:09 |
| tbarron | tell you not to use the "--cluster" arg to ceph cmd any more. | 22:09 |
| tbarron | Different clusters named "ceph" will have different fsids and if the same hardware hosts both they'll run in separate containers. | 22:10 |
| tbarron | ^^ That's my understanding of ceph community's position going forwards. | 22:10 |
| tbarron | But their apis have more tentacles and are more slimy than manila's. | 22:11 |
| tbarron | Our little zprillas were went to API finishing school despite having been called skunks by some in the ceph community. | 22:12 |
| gouthamr | :'D | 22:13 |
| vkmc | yeah, it seems an accurate outcome based on other design decisions that have been made | 22:14 |
| *** tobberydberg has quit IRC | 22:14 | |
| vkmc | "zorillas, an exotic animal that lives in africa" "oh yes, a skunk" | 22:15 |
| vkmc | </3 | 22:15 |
| *** tobberydberg has joined #openstack-manila | 22:19 | |
| tbarron | hee | 22:25 |
| *** dviroel has quit IRC | 22:55 | |
| openstackgerrit | Merged openstack/devstack-plugin-ceph master: Enable cephFS snapshots in manila https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/741275 | 22:58 |
| *** hoonetorg has quit IRC | 23:20 | |
| *** vishalmanchanda has quit IRC | 23:24 | |
| *** hoonetorg has joined #openstack-manila | 23:34 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!