| *** rhagarty has joined #openstack-manila | 00:03 | |
| *** rhagarty_ has joined #openstack-manila | 00:06 | |
| *** rhagarty has quit IRC | 00:08 | |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: [Do not merge] Testing CI network #1.2 https://review.openstack.org/281587 | 00:09 |
|---|---|---|
| *** sgotliv has quit IRC | 00:11 | |
| *** sgotliv has joined #openstack-manila | 00:11 | |
| *** baojg has quit IRC | 00:17 | |
| *** zhonghua has quit IRC | 00:21 | |
| *** huyang has quit IRC | 00:22 | |
| *** huyang has joined #openstack-manila | 00:22 | |
| *** zhonghua-lee has joined #openstack-manila | 00:24 | |
| openstackgerrit | daiki kato proposed openstack/python-manilaclient: Add missing element in functional test https://review.openstack.org/277730 | 00:33 |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: [Do not merge] Testing CI network LVM #1.4 https://review.openstack.org/281593 | 00:33 |
| *** mriedem has joined #openstack-manila | 00:36 | |
| mriedem | is it just me or does it seem like every time i try to run tox -r -e py34 it completely hoses my system | 00:38 |
| *** cknight has joined #openstack-manila | 00:45 | |
| *** cknight1 has joined #openstack-manila | 00:48 | |
| *** cknight has quit IRC | 00:50 | |
| *** jasonsb has joined #openstack-manila | 00:50 | |
| *** yangyapeng has joined #openstack-manila | 00:51 | |
| *** dcain has joined #openstack-manila | 01:05 | |
| *** yangyapeng has quit IRC | 01:06 | |
| *** dcain1 has joined #openstack-manila | 01:08 | |
| *** dcain has quit IRC | 01:12 | |
| *** mtanino has quit IRC | 01:18 | |
| bswartz | mriedem: I hope you're only hosing a VM and not your laptop | 01:19 |
| mriedem | yeah, dev vm | 01:19 |
| *** rhagarty_ has quit IRC | 01:20 | |
| *** rhagarty has joined #openstack-manila | 01:20 | |
| mriedem | my production cloud :) | 01:26 |
| *** bill_az has quit IRC | 01:28 | |
| *** yangyapeng has joined #openstack-manila | 01:31 | |
| mriedem | oh well, i'll mark it up to a fluke and push what i think is a thing that is fine on py34 | 01:33 |
| mriedem | not sure why it would not work for manila locally, but it's fine for nova locally | 01:33 |
| bswartz | oh the problem is specific to manila? | 01:33 |
| bswartz | that's not good | 01:34 |
| bswartz | the gate runs tox with py34 so I would assume that means it's not broken | 01:34 |
| bswartz | personally I'm lazy and I just run with py27 | 01:35 |
| mriedem | yeah i have a change that passes on py27 and fails on py34 so had a fix and wanted to make sure it was fine before pushing it up, | 01:36 |
| mriedem | and i've tried like 3 times to run tox -r -e py34 locally and it just hangs my dev vm | 01:36 |
| openstackgerrit | Matt Riedemann proposed openstack/manila: Cleanup deprecation warnings from using share proxy properties in API https://review.openstack.org/260186 | 01:38 |
| openstackgerrit | Matt Riedemann proposed openstack/manila: Stop proxying share_server_id through share in share.manager https://review.openstack.org/273495 | 01:38 |
| openstackgerrit | Matt Riedemann proposed openstack/manila: Get host from share['instance'] in share RPC API https://review.openstack.org/272477 | 01:38 |
| openstackgerrit | Matt Riedemann proposed openstack/manila: WIP: Remove deprecated proxy attributes from Share db model object https://review.openstack.org/260212 | 01:38 |
| openstackgerrit | Matt Riedemann proposed openstack/manila: Remove deprecated share attribute usage from manila.share.api https://review.openstack.org/272580 | 01:38 |
| *** yangyapeng has quit IRC | 01:38 | |
| mriedem | confirmed i already have python3-dev python3 installed | 01:38 |
| mriedem | anywho, ^ pushed anyway | 01:38 |
| *** yangyapeng has joined #openstack-manila | 01:38 | |
| bswartz | I'll do some experiments | 01:38 |
| *** baojg has joined #openstack-manila | 01:45 | |
| *** esker has joined #openstack-manila | 01:46 | |
| *** esker has quit IRC | 01:51 | |
| *** rhagarty has quit IRC | 01:58 | |
| *** mriedem has left #openstack-manila | 01:59 | |
| *** haomaiwa_ has joined #openstack-manila | 02:12 | |
| *** gouthamr has joined #openstack-manila | 02:15 | |
| *** gouthamr_ has joined #openstack-manila | 02:16 | |
| *** gouthamr has quit IRC | 02:20 | |
| *** houming has joined #openstack-manila | 02:35 | |
| *** a_ta has joined #openstack-manila | 02:42 | |
| *** baojg has quit IRC | 02:43 | |
| *** baojg has joined #openstack-manila | 02:47 | |
| *** a_ta has quit IRC | 02:49 | |
| *** esker has joined #openstack-manila | 02:55 | |
| *** esker has quit IRC | 03:00 | |
| *** ganso has quit IRC | 03:00 | |
| *** haomaiwa_ has quit IRC | 03:01 | |
| *** haomaiwa_ has joined #openstack-manila | 03:01 | |
| openstackgerrit | xing-yang proposed openstack/manila: Manage and unmanage snapshot https://review.openstack.org/249542 | 03:37 |
| *** baojg has quit IRC | 03:42 | |
| *** baojg has joined #openstack-manila | 03:45 | |
| *** baojg has quit IRC | 03:55 | |
| *** haomaiwa_ has quit IRC | 04:01 | |
| *** haomaiwang has joined #openstack-manila | 04:01 | |
| *** mdnadeem has joined #openstack-manila | 04:22 | |
| *** baojg has joined #openstack-manila | 04:25 | |
| mdnadeem | how to get manila-share binaries for ubuntu? | 04:39 |
| mdnadeem | It is missing in http://ubuntu-cloud.archive.canonical.com/ubuntu/dists/trusty-proposed/liberty/main/binary-amd64/Packages | 04:39 |
| mdnadeem | Any idea guys? | 04:39 |
| *** a_ta has joined #openstack-manila | 04:46 | |
| *** bill_az has joined #openstack-manila | 04:49 | |
| *** xyang1 has joined #openstack-manila | 04:50 | |
| *** a_ta has quit IRC | 04:50 | |
| openstackgerrit | Merged openstack/manila: Allow DHSS=False tests to override Tempest concurrency https://review.openstack.org/280866 | 04:56 |
| *** xyang1 has quit IRC | 04:58 | |
| *** haomaiwang has quit IRC | 05:01 | |
| *** esker has joined #openstack-manila | 05:01 | |
| *** haomaiwa_ has joined #openstack-manila | 05:01 | |
| *** esker has quit IRC | 05:05 | |
| *** esker has joined #openstack-manila | 05:21 | |
| *** esker has quit IRC | 05:26 | |
| *** bill_az has quit IRC | 05:40 | |
| *** dcain1 has quit IRC | 05:53 | |
| *** haomaiwa_ has quit IRC | 06:01 | |
| *** haomaiwang has joined #openstack-manila | 06:01 | |
| *** esker has joined #openstack-manila | 06:24 | |
| *** esker has quit IRC | 06:28 | |
| *** baojg has quit IRC | 06:47 | |
| *** a_ta has joined #openstack-manila | 06:47 | |
| *** baojg has joined #openstack-manila | 06:51 | |
| *** a_ta has quit IRC | 06:52 | |
| *** nkrinner has joined #openstack-manila | 06:58 | |
| *** lpetrut has joined #openstack-manila | 06:59 | |
| *** haomaiwang has quit IRC | 07:01 | |
| *** haomaiwa_ has joined #openstack-manila | 07:01 | |
| *** nidhimittalhada has joined #openstack-manila | 07:04 | |
| *** esker has joined #openstack-manila | 07:06 | |
| *** esker has quit IRC | 07:11 | |
| *** baojg has quit IRC | 07:13 | |
| *** baojg has joined #openstack-manila | 07:16 | |
| *** esker has joined #openstack-manila | 07:27 | |
| *** esker has quit IRC | 07:31 | |
| *** yangyape_ has joined #openstack-manila | 07:54 | |
| *** yangyap__ has joined #openstack-manila | 07:56 | |
| *** yangyapeng has quit IRC | 07:56 | |
| *** yangyape_ has quit IRC | 07:58 | |
| *** a_ta has joined #openstack-manila | 07:58 | |
| *** haomaiwa_ has quit IRC | 08:01 | |
| *** haomaiwang has joined #openstack-manila | 08:01 | |
| *** a_ta has quit IRC | 08:02 | |
| *** lpetrut has quit IRC | 08:02 | |
| *** pcaruana has joined #openstack-manila | 08:05 | |
| *** esker has joined #openstack-manila | 08:08 | |
| *** lpetrut has joined #openstack-manila | 08:10 | |
| *** esker has quit IRC | 08:13 | |
| *** lpetrut has quit IRC | 08:19 | |
| *** nidhimittalhada has quit IRC | 08:20 | |
| *** lpetrut has joined #openstack-manila | 08:43 | |
| *** rraja has joined #openstack-manila | 08:46 | |
| *** esker has joined #openstack-manila | 08:50 | |
| *** tpatzig_ has quit IRC | 08:52 | |
| *** tpatzig has joined #openstack-manila | 08:53 | |
| *** tpatzig has left #openstack-manila | 08:54 | |
| *** esker has quit IRC | 08:54 | |
| *** haomaiwang has quit IRC | 09:01 | |
| *** haomaiwang has joined #openstack-manila | 09:01 | |
| *** lpetrut has quit IRC | 09:02 | |
| *** houming has quit IRC | 09:07 | |
| *** esker has joined #openstack-manila | 09:11 | |
| *** nidhimittalhada has joined #openstack-manila | 09:13 | |
| *** esker has quit IRC | 09:15 | |
| *** yangyapeng has joined #openstack-manila | 09:19 | |
| *** yangyap__ has quit IRC | 09:21 | |
| *** yangyape_ has joined #openstack-manila | 09:22 | |
| *** yangyapeng has quit IRC | 09:25 | |
| *** ociuhandu has joined #openstack-manila | 09:29 | |
| openstackgerrit | Silvan Kaiser proposed openstack/manila: Add update_access() interface to Quobyte driver https://review.openstack.org/281136 | 09:42 |
| *** yangyape_ has quit IRC | 09:45 | |
| *** houming has joined #openstack-manila | 09:50 | |
| openstackgerrit | Abhilash Divakaran proposed openstack/manila: Add share driver for Tegile IntelliFlash Arrays https://review.openstack.org/278169 | 09:50 |
| *** gouthamr_ has quit IRC | 09:51 | |
| *** a_ta has joined #openstack-manila | 09:59 | |
| *** haomaiwang has quit IRC | 10:01 | |
| *** haomaiwang has joined #openstack-manila | 10:01 | |
| openstackgerrit | Marc Koderer proposed openstack/manila: Fix scenario tests https://review.openstack.org/280318 | 10:02 |
| *** ganso has joined #openstack-manila | 10:02 | |
| *** openstackgerrit has quit IRC | 10:02 | |
| *** openstackgerrit has joined #openstack-manila | 10:03 | |
| *** a_ta has quit IRC | 10:03 | |
| *** lpetrut has joined #openstack-manila | 10:10 | |
| *** lpetrut has quit IRC | 10:10 | |
| *** esker has joined #openstack-manila | 10:13 | |
| *** esker has quit IRC | 10:18 | |
| *** baojg has quit IRC | 10:28 | |
| *** tpsilva has joined #openstack-manila | 10:36 | |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: [Do not merge] Testing CI network LVM #1.4 https://review.openstack.org/281593 | 10:37 |
| *** esker has joined #openstack-manila | 11:00 | |
| *** marcusvrn_ has joined #openstack-manila | 11:00 | |
| *** haomaiwang has quit IRC | 11:01 | |
| *** haomaiwang has joined #openstack-manila | 11:01 | |
| *** esker has quit IRC | 11:05 | |
| tpsilva | huyang: ping | 11:07 |
| *** tobe has joined #openstack-manila | 11:11 | |
| *** esker has joined #openstack-manila | 11:15 | |
| *** esker has quit IRC | 11:20 | |
| *** ociuhandu has quit IRC | 11:22 | |
| *** mdnadeem has quit IRC | 11:22 | |
| *** erikm has quit IRC | 11:25 | |
| ganso | vponomaryov: ping | 11:29 |
| vponomaryov | ganso: pong | 11:38 |
| ganso | vponomaryov: Hello Valeriy | 11:38 |
| vponomaryov | ganso: Hello Rodrigo | 11:39 |
| ganso | vponomaryov: I tested the approach for mounting the LVM driver as we discussed yesterday, it unfortunately did not work =( http://logs.openstack.org/93/281593/2/check/gate-manila-tempest-dsvm-neutron-postgres-lvm-multibackend/a3b1f7d/logs/screen-m-dat.txt.gz | 11:39 |
| vponomaryov | ganso: try add "-v" flag for debugging | 11:40 |
| vponomaryov | ganso: also I would specify NFS version | 11:40 |
| ganso | vponomaryov: -v on the mount command? | 11:40 |
| vponomaryov | ganso: yes | 11:40 |
| vponomaryov | ganso: verbose answer | 11:40 |
| ganso | vponomaryov: ok, I will try that, thanks! | 11:40 |
| ganso | vponomaryov: there is another issue I am working on | 11:41 |
| ganso | vponomaryov: regarding scenario tests | 11:41 |
| vponomaryov | yes? | 11:41 |
| ganso | vponomaryov: after adding the admin_port, I observed the following behavior: http://logs.openstack.org/93/281593/2/check/gate-manila-tempest-dsvm-neutron-scenario-multibackend/a079be4/logs/screen-m-shr.txt.gz | 11:41 |
| ganso | vponomaryov: sometimes the service instance does not get route from service network | 11:41 |
| vponomaryov | line? | 11:42 |
| ganso | vponomaryov: http://logs.openstack.org/93/281593/2/check/gate-manila-tempest-dsvm-neutron-scenario-multibackend/a079be4/logs/screen-m-shr.txt.gz#_2016-02-18_11_15_05_064 | 11:42 |
| ganso | vponomaryov: I created a debug patch that extends the admin_port patch | 11:42 |
| ganso | vponomaryov: the patch enforces the service_port to the eth0 in the instance | 11:42 |
| ganso | vponomaryov: and it also creates another network to use as admin instead of previously created "private" | 11:43 |
| ganso | vponomaryov: this new network has "--no-gateway" | 11:43 |
| ganso | vponomaryov: this was an attempt to not conflict with service_instance route assignment | 11:43 |
| ganso | vponomaryov: but now it is clear that even though there is no conflict, there is also no route from the service_network... | 11:44 |
| ganso | vponomaryov: it mostly happens in the 2nd service instance created in the job onwards | 11:44 |
| ganso | vponomaryov: I can recreate this in my local lab, but it is random, sometimes it works, sometimes it does not | 11:44 |
| ganso | vponomaryov: like, running scenario test 10x repeadtly in my local lab, it passed only twice | 11:45 |
| *** lpetrut has joined #openstack-manila | 11:45 | |
| *** houming has quit IRC | 11:45 | |
| vponomaryov | ganso: IP 104.130.13.28 ? | 11:47 |
| ganso | vponomaryov: what's that? | 11:48 |
| vponomaryov | ganso: it is IP you try ping and it is out of local network range | 11:48 |
| ganso | vponomaryov: I ping it twice | 11:49 |
| ganso | vponomaryov: first, when there is route, it works | 11:49 |
| ganso | vponomaryov: when there is not, it does not work | 11:49 |
| vponomaryov | ganso: so, you wonder why route is absent in second case? | 11:49 |
| ganso | vponomaryov: yes | 11:50 |
| ganso | vponomaryov: I supposed all service instances are created the same way, but something is wrong | 11:50 |
| ganso | vponomaryov: this could be what's impacting the scenario tests stability | 11:50 |
| ganso | vponomaryov: mkoderer 's patch that disabled migration still did not fix scenario tests | 11:51 |
| ganso | vponomaryov: looking at nova/neutron logs, it seems like it tries to assign the route to the VM, but the VM does not get it assigned, and I could not find any errors from nova/neutron or dmesg (in my local lab) | 11:54 |
| *** rraja has quit IRC | 11:54 | |
| vponomaryov | ganso: port is up? | 11:56 |
| ganso | vponomaryov: yes | 11:56 |
| vponomaryov | ganso: always? | 11:56 |
| ganso | vponomaryov: every time I checked, yes | 11:56 |
| ganso | vponomaryov: http://logs.openstack.org/93/281593/2/check/gate-manila-tempest-dsvm-neutron-scenario-multibackend/a079be4/logs/screen-m-shr.txt.gz#_2016-02-18_11_00_26_258 | 11:57 |
| ganso | vponomaryov: easy to search for "manila-admin-share" | 11:57 |
| vponomaryov | ganso: and you sure your local devices code does not break anything? | 11:58 |
| vponomaryov | ganso: that you added? | 11:58 |
| vponomaryov | ganso: for example, cleanup? | 11:59 |
| vponomaryov | ganso: does second call break first one's routes? | 11:59 |
| ganso | vponomaryov: the problem happens before any cleanup is done | 11:59 |
| ganso | vponomaryov: oh you mean cleanup of routes on the host/ | 11:59 |
| vponomaryov | ganso: yes | 12:00 |
| vponomaryov | ganso: linux_lib code | 12:00 |
| vponomaryov | ganso: try create three-four Vms in local lab | 12:00 |
| vponomaryov | and connect to each of them | 12:00 |
| vponomaryov | from host | 12:00 |
| vponomaryov | deleting one of them in the middle | 12:00 |
| ganso | vponomaryov: it is performed on the host, the route prints are inside the service instance. | 12:00 |
| *** haomaiwang has quit IRC | 12:01 | |
| vponomaryov | host can lost connection | 12:01 |
| *** haomaiwa_ has joined #openstack-manila | 12:01 | |
| vponomaryov | if routes are broken by logic of "linux_lib" | 12:01 |
| vponomaryov | ganso: so, I would try first mentioned case | 12:01 |
| ganso | vponomaryov: ok, I will try that now, I certainly tried creating 3 VMs, they are all accessible from the host, but they have route problems inside that result in VMs not being able to connect to them. I will try deleting one and see what happens | 12:01 |
| vponomaryov | first always ok, right? | 12:02 |
| ganso | vponomaryov: first is always ok | 12:02 |
| vponomaryov | what is first - delete, second? | 12:02 |
| vponomaryov | s/is/if/ | 12:02 |
| vponomaryov | looks like it is true | 12:03 |
| vponomaryov | works only first in list | 12:03 |
| ganso | vponomaryov: haven't tried deleting before creating another one. I always create a second one right away | 12:03 |
| ganso | vponomaryov: that's one more test I will do | 12:03 |
| ganso | vponomaryov: thanks for the suggestion | 12:04 |
| mkoderer | ganso: the last run was green | 12:09 |
| mkoderer | https://jenkins03.openstack.org/job/gate-manila-tempest-dsvm-neutron-scenario-multibackend/450/console | 12:09 |
| mkoderer | ganso: but it's much too flaky | 12:09 |
| *** nidhimittalhada has quit IRC | 12:10 | |
| ganso | mkoderer: yes. it is :( | 12:11 |
| ganso | mkoderer: it is passble now at least | 12:11 |
| *** erlon has joined #openstack-manila | 12:11 | |
| ganso | mkoderer: but I would like to dig into the root cause | 12:11 |
| mkoderer | ganso: +1 | 12:12 |
| mkoderer | ganso: I would really like to have them stable and voting again | 12:12 |
| openstackgerrit | Valeriy Ponomaryov proposed openstack/manila: Add ZFSonLinux share driver https://review.openstack.org/277192 | 12:14 |
| *** ociuhandu has joined #openstack-manila | 12:23 | |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Add new Share migration API functions https://review.openstack.org/259642 | 12:27 |
| *** JoseMello has joined #openstack-manila | 12:28 | |
| *** lpabon has joined #openstack-manila | 12:34 | |
| *** yangyapeng has joined #openstack-manila | 12:46 | |
| *** haomaiwa_ has quit IRC | 12:52 | |
| *** timcl has joined #openstack-manila | 12:52 | |
| *** a_ta has joined #openstack-manila | 12:54 | |
| openstackgerrit | Silvan Kaiser proposed openstack/manila: Add update_access() interface to Quobyte driver https://review.openstack.org/281136 | 12:55 |
| *** a_ta has quit IRC | 12:58 | |
| openstackgerrit | Silvan Kaiser proposed openstack/manila: Add update_access() interface to Quobyte driver https://review.openstack.org/281136 | 12:59 |
| *** limamauricio has quit IRC | 13:02 | |
| *** haomaiwang has joined #openstack-manila | 13:22 | |
| *** nkrinner has quit IRC | 13:28 | |
| *** nkrinner has joined #openstack-manila | 13:28 | |
| *** xyang1 has joined #openstack-manila | 13:47 | |
| *** xyang_ has joined #openstack-manila | 13:48 | |
| *** martyturner has joined #openstack-manila | 13:56 | |
| *** haomaiwang has quit IRC | 14:01 | |
| *** haomaiwang has joined #openstack-manila | 14:01 | |
| openstackgerrit | Thomas Bechtold proposed openstack/manila: Check for device node availability before mkfs https://review.openstack.org/281849 | 14:03 |
| mkoderer | ganso: do you think we should merge https://review.openstack.org/#/c/280318/ to at least fix the obvious issues? | 14:03 |
| xyang_ | bswartz: regarding https://review.openstack.org/#/c/249542/, cknight has a question on why provider_location is in share_snapshots, can we discuss about it? | 14:12 |
| *** akerr has joined #openstack-manila | 14:17 | |
| openstackgerrit | Valeriy Ponomaryov proposed openstack/manila: Add ZFSonLinux share driver https://review.openstack.org/277192 | 14:17 |
| cknight1 | xyang1, bswartz: Good morning, y'all. Since provider_location is very specific to where a snapshot is kept on a given backend, I didn't understand why that field wasn't on SnapshotInstance instead of Snapshot. But the good news is that the code is in pretty good shape now. The interface appears clean & consistent with share manage, and it worked well in the NetApp driver. We hope to try the Tempest part against our driver today. | 14:19 |
| xyang_ | cknight1: thanks for testing this. | 14:21 |
| vponomaryov | xyang1, cknight: I would agree that this attr - "provide_location" should be set exactly to "SnapshotInstance" model | 14:21 |
| vponomaryov | xyang_: it is instance-specific | 14:22 |
| xyang_ | cknight1, vponomaryov, bswartz: if we change this to SnapshotInstance, I cannot get it via snapshot_get (with admin context) any more. bswartz has suggested to return provider_location by unmanage_snapshot, I can try that if everyone agrees | 14:23 |
| vponomaryov | xyang_: +1 for return value | 14:24 |
| vponomaryov | xyang_: list of values for each of instances | 14:24 |
| * vponomaryov will be back in 20-30 minutes | 14:24 | |
| xyang_ | vponomaryov: [{'snapshot_instance': {'id': xxx, 'provider_location': xxx}}, {'id': xxx}]? | 14:25 |
| cknight1 | xyang1: I wouldn't do a bunch of coding until we get agreement with Ben. I did like getting the provider location via the REST interface, but it's a problem if there aren't snapshot instance list APIs. | 14:26 |
| xyang_ | cknight1: yes, we don't have that API yet. | 14:26 |
| cknight1 | xyang1: OK, so perhaps you return the value from unmanage, *and* you store it in the snapshot instance table, so when we add the API it's already there for us. But let's discuss with Ben when he's available. | 14:28 |
| xyang_ | cknight1: bswartz has suggested to return provider_location from unmanage api. sure, let's wait for him | 14:28 |
| *** dcain has joined #openstack-manila | 14:28 | |
| xyang_ | cknight1: I think he is busy in cinder | 14:28 |
| cknight1 | xyang1: yes, arguing about microversions. glad we're past that in manila! | 14:29 |
| xyang_ | cknight1: +1:) | 14:29 |
| cknight1 | xyang1: glad we don't have much arguing in manila. | 14:29 |
| *** gouthamr has joined #openstack-manila | 14:30 | |
| jcsp | ganso: is the access rule status field still meaningful after the update_access change? it seems like errors are now handled by doing share_instance_update_access_status | 14:30 |
| jcsp | which seems like a global flag per share rather than a status per access rule | 14:30 |
| *** eharney has quit IRC | 14:33 | |
| mkoderer | cknight1: hi, may I ask you again a netapp driver question ;) | 14:34 |
| cknight1 | mkoderer: ok. i'm just about to leave for the office, but I'll try either way. | 14:34 |
| *** mtanino has joined #openstack-manila | 14:35 | |
| *** cknight1 has quit IRC | 14:35 | |
| mkoderer | cknight1: ok let's wait until your in the office :) | 14:37 |
| *** porrua has joined #openstack-manila | 14:39 | |
| ganso | jcsp: hi | 14:39 |
| ganso | jcsp: it is meaningful only for backwards compatibility | 14:40 |
| jcsp | so it'll always say "active"? | 14:40 |
| ganso | jcsp: if you can check share_instance_update_access_status it will be much more accurate | 14:40 |
| ganso | jcsp: no | 14:40 |
| ganso | jcsp: it will say the same thing as share_instance_update_access_status | 14:40 |
| jcsp | oh, o | 14:40 |
| jcsp | *ok | 14:40 |
| jcsp | thanks | 14:41 |
| ganso | jcsp: like, if you have 10 access rules, and you add a new one, the status field of all rules will go to "new" at the same time | 14:41 |
| ganso | jcsp: and then all back to active once the rule is applied | 14:41 |
| ganso | jcsp: the status cannot be individual anymore | 14:41 |
| jcsp | is there a plan for restoring the ability for user to see which rule caused an error? | 14:43 |
| ganso | jcsp: I am not sure about that. Since in the future we plan to allow rules to be added in groups | 14:43 |
| ganso | mkoderer: hi | 14:44 |
| ganso | mkoderer: I am still investigating that... I will have a position about this soon | 14:44 |
| *** dustins has joined #openstack-manila | 14:47 | |
| *** bill_az has joined #openstack-manila | 14:48 | |
| vponomaryov | xyang_: vponomaryov: [{'snapshot_instance': {'id': xxx, 'provider_location': xxx}}, {'id': yyy, 'provider_location': yyy}]? - yes | 14:49 |
| *** eharney has joined #openstack-manila | 14:50 | |
| vponomaryov | xyang_: not counting syntax errors )) | 14:50 |
| xyang_ | vponomaryov: yes, type quick | 14:50 |
| bswartz | xyang_: ping | 14:50 |
| bswartz | xyang_: cknight seems to have dropped offline | 14:50 |
| xyang_ | bswartz: hi, did you see the conversation | 14:50 |
| *** zhonghua has joined #openstack-manila | 14:50 | |
| bswartz | I didn't want to return the string from unmanage | 14:51 |
| xyang_ | bswartz: so right now provider_location is in share_snapshots | 14:51 |
| bswartz | since you had the idea to return it from the get snapshot API | 14:51 |
| *** rraja has joined #openstack-manila | 14:51 | |
| bswartz | as long as it's admin-only, returning it from the get snapshot API is fine IMO | 14:51 |
| bswartz | I originally wanted to return it from unmanage but you convinced me your way was better | 14:51 |
| xyang_ | bswartz: so cknight's question is why provider location is in share_snapshots not in share_snapshot_instances | 14:52 |
| bswartz | oh | 14:52 |
| bswartz | that's a different question | 14:52 |
| xyang_ | bswartz: that is related | 14:52 |
| xyang_ | bswartz: so we talked about adding provider_location to share_snapshots earlier, that is how I implemented it, but if we move it to share_snapshot_instances, we need to talk about how to retrieve it again | 14:53 |
| *** zhonghua-lee has quit IRC | 14:53 | |
| xyang_ | bswartz: so should provider_location be in share_snapshots? | 14:53 |
| openstackgerrit | Marc Koderer proposed openstack/manila: Fix scenario tests https://review.openstack.org/280318 | 14:54 |
| bswartz | my instinct says it should be on the snapshot | 14:55 |
| bswartz | but I haven't thought through the implications | 14:55 |
| bswartz | like what happens if you migrate a share | 14:55 |
| bswartz | or replicate | 14:55 |
| *** martyturner has quit IRC | 14:57 | |
| xyang_ | bswartz: yes, it is getting complicated if we migrate or replicate. | 14:57 |
| *** tobe has quit IRC | 14:57 | |
| bswartz | xyang_: so currently, migrations delete all your snapshots | 14:58 |
| *** cknight has joined #openstack-manila | 14:58 | |
| bswartz | which means we can ignore that case for the most part | 14:58 |
| bswartz | and replication isn't supported with manage/unmanage (you should get an error trying to manage a replicated share type) | 14:59 |
| *** martyturner has joined #openstack-manila | 14:59 | |
| *** haomaiwang has quit IRC | 15:01 | |
| *** a_ta has joined #openstack-manila | 15:01 | |
| *** haomaiwang has joined #openstack-manila | 15:01 | |
| *** markstur_ has joined #openstack-manila | 15:02 | |
| *** rhagarty has joined #openstack-manila | 15:03 | |
| *** markstur_ has left #openstack-manila | 15:03 | |
| *** merooney has joined #openstack-manila | 15:03 | |
| *** mriedem has joined #openstack-manila | 15:04 | |
| mriedem | has anyone seen this before? http://logs.openstack.org/86/260186/15/check/gate-manila-tempest-dsvm-neutron/2877658/logs/screen-m-shr.txt.gz#_2016-02-18_02_07_15_218 | 15:05 |
| *** dcain has quit IRC | 15:05 | |
| vponomaryov | mriedem: yes | 15:05 |
| mriedem | looks like it's been around awhile: | 15:06 |
| mriedem | http://logstash.openstack.org/#dashboard/file/logstash.json?query=message%3A%5C%22Server%20Information%20in%20exception%20can%20not%20be%20written%20to%20db%20%3A%20Invalid%20server_details%20'None'%5C%22%20AND%20tags%3A%5C%22screen-m-shr.txt%5C%22 | 15:06 |
| mriedem | http://logstash.openstack.org/#dashboard/file/logstash.json?query=message%3A%5C%22Server%20Information%20in%20exception%20can%20not%20be%20written%20to%20db%20%3A%20Invalid%20server_details%20'None'%5C%22%20AND%20tags%3A%5C%22screen-m-shr.txt%5C%22&from=7d | 15:06 |
| *** dcain has joined #openstack-manila | 15:06 | |
| mriedem | was trying to sort out why this happened http://logs.openstack.org/86/260186/15/check/gate-manila-tempest-dsvm-neutron/2877658/console.html#_2016-02-18_02_37_12_427 | 15:06 |
| *** esker has joined #openstack-manila | 15:07 | |
| *** esker has quit IRC | 15:07 | |
| mriedem | There isn't much for details in the share logs for the cg, just 4 entries for that uuid | 15:07 |
| mriedem | creating and then deleting basically | 15:07 |
| mriedem | no explanation as to why it went into error state | 15:07 |
| *** esker has joined #openstack-manila | 15:07 | |
| vponomaryov | mriedem: ServiceInstanceException: SSH connection has not been established to 10.254.0.5 in 300s. Giving up. | 15:09 |
| vponomaryov | mriedem: Nova VM problems | 15:09 |
| bswartz | mriedem: last night I ran tox -e py34 and it worked with no problems on my VM (ubuntu trusty) | 15:10 |
| mriedem | bswartz: huh, ok | 15:15 |
| openstackgerrit | Matt Riedemann proposed openstack/manila: Make consistency group timeout exception message more robust https://review.openstack.org/281904 | 15:17 |
| bswartz | I did have to apt-get install python3-dev, but that was it | 15:19 |
| mriedem | yeah i checked that, had it installed | 15:19 |
| mriedem | must be a fluke | 15:19 |
| mriedem | wrt this http://logs.openstack.org/86/260186/15/check/gate-manila-tempest-dsvm-neutron/2877658/logs/screen-m-shr.txt.gz#_2016-02-18_02_07_15_218 | 15:20 |
| mriedem | is that a programming error if the exception doesn't have a detail_data key with a server_details key in that? | 15:20 |
| mriedem | otherwise i'm not sure what an operator / admin seeing that warning is supposed to do about it | 15:21 |
| mriedem | and w/o a stacktrace, i'm not sure how a developer knows where it came from | 15:21 |
| bswartz | vponomaryov: ^ | 15:23 |
| *** haomaiwang has quit IRC | 15:25 | |
| openstackgerrit | ting wang proposed openstack/manila: Use ostestr to run unit test https://review.openstack.org/281910 | 15:30 |
| *** bill_az has quit IRC | 15:30 | |
| *** martyturner has quit IRC | 15:34 | |
| bswartz | jcsp: over here | 15:34 |
| xyang_ | bswartz, cknight: can we continue the discussion | 15:34 |
| jcsp | bswartz: *waves* | 15:34 |
| cknight | xyang1: I'm here if bswartz is. | 15:35 |
| xyang_ | bswartz: you need a clone:) | 15:35 |
| cknight | xyang1: I think we all do. | 15:35 |
| xyang_ | cknight: bswartz was just talking about provider_location in snapshots table before the meeting started | 15:35 |
| vponomaryov | mriedem: according to error "Server Information in exception can not be written to db : Invalid server_details 'None' " feel free to file a bug. It is definitly bug that should be fixed | 15:36 |
| mriedem | vponomaryov: ok, will do | 15:37 |
| ganso | mkoderer: ping | 15:39 |
| bswartz | sorry my phone rang | 15:40 |
| bswartz | I really do need a clone | 15:40 |
| xyang_ | :) | 15:40 |
| *** tobe has joined #openstack-manila | 15:40 | |
| bswartz | jcsp: regarding the update access method, it was designed in such a way that, when we're just adding or removing rules, those rules are passed in so you can do an optimized approach | 15:41 |
| bswartz | that design was based on the concerns you raised in tokyo | 15:41 |
| bswartz | so I don't see any downside for you to implement it | 15:41 |
| bswartz | if it's a question of timing, and you want to do a separate patch for it, I think we could live with that | 15:42 |
| bswartz | it would be better to get it into mitaka than to ship mitaka with the old interface | 15:42 |
| jcsp | the timing is the main thing. If we have to do it right this second then the implementation will just be effectively a copy of the fallback code, because the interface to Ceph doesn't have a way to list existing rules | 15:43 |
| bswartz | jcsp: does it have a way to "replace all" rules? | 15:43 |
| jcsp | so the design enables us to do a fallback, but it remains the case that there's no actual upside/benefit to this when using ceph. We still sit there in a loop doing allow calls. | 15:43 |
| jcsp | nope. | 15:43 |
| bswartz | is there even a "delete all"? | 15:43 |
| jcsp | nope. | 15:44 |
| bswartz | so.... | 15:44 |
| jcsp | it's just not a bulk interface at all. | 15:44 |
| bswartz | what do you do when you don't know what the existing rules are and you want to change them>? | 15:44 |
| jcsp | in the lower level ceph interface (that our CLI uses) you can list rules, and then go do your one-by-one operations on them. | 15:44 |
| jcsp | we built the manila driver on top of an abstraction that gives a "volume" concept | 15:45 |
| jcsp | and that abstraction doesn't have the listing part (yet) | 15:45 |
| bswartz | so is that API omission something that will be fixed? | 15:45 |
| jcsp | but, to be clear, even when we add that, we're just doing it to make manila happy. It's not helping anyone, because the actual operations are still all one by one | 15:45 |
| jcsp | yes, I expect we'll extend the API to enable Manila's update_access. | 15:46 |
| bswartz | so I'm envisioning a future where manila access rules are defined in terms of stuff like nova instances and neutron security groups | 15:46 |
| *** haomaiwang has joined #openstack-manila | 15:46 | |
| bswartz | in that future, the list of IPs might change due to external opterations | 15:46 |
| *** yangyapeng has quit IRC | 15:47 | |
| bswartz | and computing the diff between the old ones and the new ones might be hard or impossible for manila to do | 15:47 |
| bswartz | manila will just know that the set of IPs that should gave access to a share is different than it used to be and it will have the new list | 15:48 |
| jcsp | You're talking about IPs here. None of that matters to Ceph. | 15:48 |
| jcsp | I get that some drivers need it | 15:49 |
| bswartz | jcsp: what about CephNFS? | 15:49 |
| bswartz | that was something you mentioned you planned to support in tokyo | 15:49 |
| jcsp | if/when someone writes a driver that layers a cluster of NFS servers on top of CephFS, they'll be configuring the access rules within the NFS servers. The interface for that will depend on what NFS server is in use. | 15:50 |
| jcsp | e.g. if someone is using ganesha gateways it'll depend on how well they can go query their ganesha config or bulk-overwrite their ganesha config | 15:50 |
| bswartz | okay so your current limitation only applies when you're dealing with ceph credentials and not IPs | 15:51 |
| bswartz | I understand | 15:51 |
| *** haomaiwang has quit IRC | 15:51 | |
| bswartz | in that case you won't really benefit from the new interface | 15:51 |
| *** bill_az has joined #openstack-manila | 15:51 | |
| bswartz | but we still want to remove the old interface ASAP | 15:51 |
| jcsp | so as it happens I've already written some code that just effectively embeds the fallback into the ceph driver | 15:52 |
| bswartz | I think you can implement update_access as a loop that calls the existing allow/deny functions | 15:52 |
| jcsp | right, so doing that is pointless as long as the fallback is in place in Manila, but if folks want to remove the fallback then there is a point to me doing it. | 15:52 |
| jcsp | if folks are okay with this review just having the obvious loop instead of a "real" implementation then I can just stick that in here | 15:53 |
| bswartz | yes | 15:53 |
| openstackgerrit | Matt Riedemann proposed openstack/manila: WIP: pass catalog info when creating novaclient for admin context https://review.openstack.org/277140 | 15:53 |
| bswartz | I'm okay with that | 15:53 |
| bswartz | xyang_, xyang1: you still around? | 15:54 |
| jcsp | OK, thanks | 15:54 |
| xyang_ | bswartz: yes | 15:54 |
| xyang_ | cknight: ^ | 15:54 |
| bswartz | so where we left off before the meeting | 15:54 |
| xyang_ | bswartz: you were talking about keeping provider_location in snapshots table, whether it will affect migration and replication | 15:55 |
| bswartz | Is there a case where there could be 2 provider locations? | 15:55 |
| bswartz | for 2 instances? | 15:55 |
| xyang_ | bswartz: it depends on how we use provider location | 15:55 |
| xyang_ | bswartz: it is a string | 15:56 |
| bswartz | migration should wipe out the snapshots, so while there would be 2 share instances, there wouldn't be 2 snapshot instances | 15:56 |
| xyang_ | ok, so we don't have to worry about migration now | 15:57 |
| *** dcain has quit IRC | 15:57 | |
| bswartz | so replicated shares will have multiple snapshot instances | 15:58 |
| bswartz | and each snapshot instance could have a different provider location | 15:58 |
| bswartz | however we don't plan to support manage/unmanage of shares that are replicated | 15:59 |
| bswartz | for 2 reasions: (1) it's a ton of work (2) there no clear use case | 15:59 |
| *** yangyapeng has joined #openstack-manila | 15:59 | |
| *** krotscheck_dcm is now known as krotscheck | 16:00 | |
| xyang_ | bswartz: knight: so keeping provider_location in share_snapshots table? | 16:00 |
| *** timcl has quit IRC | 16:01 | |
| bswartz | well we could put it in the instances table | 16:01 |
| xyang_ | bswartz: if we do that, we need to discuss how to retrieve it next | 16:02 |
| bswartz | but still return it in the snapshots API (just return the value from the "active" instance) | 16:02 |
| xyang_ | bswartz: right now I use get_snapshot with admin context to retrieve provider_location from snapshots table | 16:02 |
| xyang_ | bswartz: if it is no longer in snapshots table, I cannot use get_snapshot any more | 16:02 |
| bswartz | why can't you access the instances when creating the snapshot view? | 16:03 |
| bswartz | aren't they part of the sqlalchemy model? | 16:03 |
| bswartz | cknight: where did you go? | 16:03 |
| xyang_ | bswartz: that is not exposed in the view | 16:03 |
| xyang_ | bswartz: I can add that code, but not sure if that is what we wanted | 16:03 |
| *** zigo has quit IRC | 16:03 | |
| bswartz | okay so we're in agreement | 16:04 |
| xyang_ | bswartz: yes, it is in models, but not in views | 16:04 |
| xyang_ | bswartz: ? | 16:04 |
| bswartz | I'm not sure if that's what we want either -- that's what we need to decide | 16:04 |
| bswartz | however we agree that it's possible to do this in the view code | 16:04 |
| cknight | bswartz: here now | 16:04 |
| xyang_ | bswartz: get_snapshot will get provider_location from snapshot_instances table? | 16:04 |
| xyang_ | bswartz: with admin context | 16:04 |
| *** zigo has joined #openstack-manila | 16:05 | |
| bswartz | cknight: the proposal is to store provider_location in the snapshot instance table, but to only return 1 provider location in the view for the snapshot | 16:05 |
| cknight | bswartz: And I'm OK with that. | 16:05 |
| *** rraja has quit IRC | 16:05 | |
| xyang_ | bswartz, cknight, alright, I can make that change | 16:05 |
| bswartz | cknight: we believe that won't cause any problems because manage should work with a replicated share and there are no other cases where you'll have 2 instances of a snapshot | 16:05 |
| bswartz | manage should *NOT* work with a replicated share | 16:05 |
| * bswartz fails at typing | 16:06 | |
| cknight | bswartz: agreed | 16:06 |
| xyang_ | vponomaryov: are you ok | 16:06 |
| *** gregsfortytwo1 has quit IRC | 16:06 | |
| *** mriedem has left #openstack-manila | 16:07 | |
| vponomaryov | ok | 16:08 |
| xyang_ | great, I'll update the patch | 16:09 |
| *** martyturner has joined #openstack-manila | 16:11 | |
| *** dcain has joined #openstack-manila | 16:15 | |
| *** pcaruana has quit IRC | 16:15 | |
| *** nkrinner has quit IRC | 16:16 | |
| *** yangyapeng has quit IRC | 16:17 | |
| ganso | bswartz: LVM is still a bit unstable http://logs.openstack.org/81/256281/24/check/gate-manila-tempest-dsvm-neutron-postgres-lvm-multibackend/cf7ac6d/logs/screen-m-shr.txt.gz?level=ERROR | 16:17 |
| ganso | bswartz: in the same patch, it had previously passed, now it did not pass, I am going to recheck again | 16:17 |
| *** gregsfortytwo has joined #openstack-manila | 16:19 | |
| *** bill_az has quit IRC | 16:23 | |
| bswartz | ganso: https://review.openstack.org/#/c/276560/ | 16:24 |
| openstackgerrit | Ben Swartzlander proposed openstack/manila: Change tempest concurrency to 1 for LVM driver https://review.openstack.org/276560 | 16:24 |
| bswartz | ganso: based on these tests I suspect the driver has concurrency problem | 16:24 |
| bswartz | it's possible that the concurrently problem lies in the tempest tests, but more likely it's in the driver | 16:25 |
| ganso | bswartz: since LVM driver does not use neutron, I would say the concurrency problem is in the driver as well | 16:25 |
| ganso | bswartz: I am starting to think about setting concurrency 1 for scenario tests, I believe the problems are related to tempest + neutron | 16:26 |
| bswartz | ganso: it's going to require some serious debugging to figure out for sure | 16:28 |
| *** timcl has joined #openstack-manila | 16:31 | |
| *** haomaiwang has joined #openstack-manila | 16:32 | |
| openstackgerrit | Valeriy Ponomaryov proposed openstack/manila: Add ZFSonLinux share driver https://review.openstack.org/277192 | 16:34 |
| ganso | bswartz: mkoderer and I are invested in debugging in order to fix scenario tests | 16:34 |
| ganso | bswartz: we have a patch in gerrit to fix it | 16:34 |
| ganso | bswartz: without this patch, scenario tests never pass | 16:34 |
| ganso | bswartz: with the patch, it rarely does... but it is passable | 16:35 |
| ganso | bswartz: I am not happy with it | 16:35 |
| bswartz | ganso: link? | 16:35 |
| ganso | bswartz: https://review.openstack.org/#/c/280318/ | 16:36 |
| ganso | bswartz: in my local lab, concurrency=1 is not enough to fix it | 16:36 |
| ganso | bswartz: it is still unstable | 16:36 |
| bswartz | ganso: btw thanks for working on that -- better test coverage is hugely important and ususally we ignore it until after feature freeze :-[ | 16:36 |
| ganso | bswartz: yup, we ignored it for update_access() patch, which broke it | 16:36 |
| ganso | bswartz: there is too much effort to fix the old migration code, now that it is refactored, and it would be for a job that rarely passes | 16:37 |
| ganso | bswartz: so I already fixed migration scenario test in my migration patches | 16:38 |
| ganso | bswartz: but the root cause of instability still prevents it for passing consistently | 16:38 |
| openstackgerrit | Clinton Knight proposed openstack/manila: Support export location metadata in NetApp cDOT drivers https://review.openstack.org/280807 | 16:44 |
| openstackgerrit | Clinton Knight proposed openstack/manila: Admin networks in NetApp cDOT multi-SVM driver https://review.openstack.org/281950 | 16:44 |
| ganso | cknight: have you tested migration in your netapp admin_network patch? | 16:45 |
| cknight | ganso: no, haven't had time. but that patch is specifically so that we *can* test migration. | 16:46 |
| ganso | cknight: =) | 16:46 |
| cknight | ganso: you'd be welcome to review. I've started reading the patches you asked me about, but I've been mostly focused on the deadline today. | 16:47 |
| ganso | cknight: sure. I am focused right now in the last attempts to fix scenario tests, soon I will be done with that and take a look at your patch | 16:48 |
| *** xyang_ has quit IRC | 16:48 | |
| *** dcain has quit IRC | 16:49 | |
| cknight | ganso: fair enough, thanks. | 16:49 |
| cknight | ganso: after today, it's all reviewing & merging. | 16:50 |
| ganso | cknight: yup, looking forward to it | 16:50 |
| cknight | ganso: Here's a query that can help with rechecks: https://review.openstack.org/#/q/project:%22%255Eopenstack/python-manilaclient%257Copenstack/manila%257Copenstack/manila-ui%22++label:Code-Review%253D%252B2++label:Workflow%253D%252B1++(label:Verified-1+OR+label:Verified-2)++status:open++is:mergeable | 16:50 |
| ganso | cknight: thanks! that is going to be very useful! | 16:51 |
| *** dcain has joined #openstack-manila | 16:59 | |
| *** haomaiwang has quit IRC | 17:01 | |
| *** haomaiwa_ has joined #openstack-manila | 17:01 | |
| *** esker has quit IRC | 17:03 | |
| *** esker has joined #openstack-manila | 17:03 | |
| *** rhagarty_ has joined #openstack-manila | 17:08 | |
| *** rhagarty has quit IRC | 17:10 | |
| *** esker has quit IRC | 17:11 | |
| *** dencaval has quit IRC | 17:12 | |
| *** tobe has quit IRC | 17:13 | |
| openstackgerrit | Tiago Pasqualini da Silva proposed openstack/manila: Add separate process for Data Service copy https://review.openstack.org/279164 | 17:14 |
| *** a_ta has quit IRC | 17:15 | |
| *** a_ta has joined #openstack-manila | 17:16 | |
| *** lgreg has joined #openstack-manila | 17:20 | |
| *** a_ta has quit IRC | 17:20 | |
| *** a_ta has joined #openstack-manila | 17:22 | |
| *** dencaval has joined #openstack-manila | 17:22 | |
| *** krotscheck is now known as krotscheck_dr | 17:27 | |
| bswartz | ganso: akerr filed a bug related to tests, and here is a fix: https://review.openstack.org/#/c/281534/ | 17:50 |
| bswartz | I'm not sure if this affects the cases you've been seeing | 17:51 |
| *** cknight has quit IRC | 17:53 | |
| *** timcl has quit IRC | 17:53 | |
| *** limamauricio has joined #openstack-manila | 17:54 | |
| ganso | bswartz: unrelated | 17:55 |
| dencaval | Guys, is there a way to set the manila service network to use a physical nic or should the manila service be connected to a tenant network? | 17:55 |
| ganso | bswartz: thanks for letting me know | 17:55 |
| limamauricio | I'm trying to create a share but now i have this error -> ERROR oslo_db.sqlalchemy.exc_filters IntegrityError: (1048, u"Column 'value' cannot be null") | 17:56 |
| limamauricio | more details -> http://pastebin.com/8VLAx8RD | 17:56 |
| bswartz | dencaval: it depends on the network plugin and driver | 17:59 |
| dencaval | dencaval I'm using generic driver with DHSS enabled and linux bridge plugin. | 18:00 |
| dencaval | bswartz, I'm using generic driver with DHSS enabled and linux bridge plugin. | 18:00 |
| *** haomaiwa_ has quit IRC | 18:01 | |
| *** haomaiwang has joined #openstack-manila | 18:01 | |
| *** haomaiwang is now known as 16WAAAVR4 | 18:01 | |
| *** timcl has joined #openstack-manila | 18:01 | |
| bswartz | tenant network in that case | 18:02 |
| bswartz | it should be a tenant network in most cases | 18:02 |
| dencaval | bswartz, Actually I having a error that the share node can't not ssh the share server at compute node. | 18:02 |
| bswartz | however we do support the concept of "flat networking" which basically means Manila doesn't do anything and you're on your own | 18:02 |
| *** lpetrut has quit IRC | 18:02 | |
| bswartz | dencaval: you might have hit our least favorite neutron bug | 18:03 |
| dencaval | bswartz, How so? I think the default manila service network is not reachable by the my share node. | 18:04 |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: [Do not merge] Testing CI Scenario tests https://review.openstack.org/281993 | 18:05 |
| bswartz | dencaval: oh you may be right | 18:05 |
| bswartz | there is another limitation with the generic driver | 18:05 |
| bswartz | the generic driver requires that the neutron agent service is running on the same node where the manila-share service is running | 18:06 |
| bswartz | it's a requirement for only the generic driver | 18:06 |
| dencaval | bswartz, hmm | 18:07 |
| bswartz | dencaval: are you running them on the same node or separate? | 18:07 |
| bswartz | dencaval: for the purposes of the install guide, the LVM driver would provide a less complex example setup for new users | 18:07 |
| dencaval | bswartz, deployment separate: controller node, compute node and share node (without nova and neutron) | 18:07 |
| bswartz | and it would be more similar if flavor to the cinder install guide and its requirements | 18:08 |
| bswartz | s/if/in/ | 18:08 |
| *** a_ta has quit IRC | 18:08 | |
| *** a_ta has joined #openstack-manila | 18:09 | |
| vponomaryov | limamauricio: "router_id" is None for some reason | 18:10 |
| limamauricio | vponomaryov: i know, but i set it | 18:11 |
| vponomaryov | limamauricio: what do you mean you set it? | 18:11 |
| vponomaryov | limamauricio: it should be autopicked up based on network | 18:12 |
| limamauricio | my bad | 18:12 |
| limamauricio | i set sub-net id | 18:12 |
| vponomaryov | limamauricio: also, is your network switched to public router? | 18:12 |
| vponomaryov | in default configuration of Generic driver it is required | 18:13 |
| *** a_ta has quit IRC | 18:14 | |
| *** fthiagogv_ has joined #openstack-manila | 18:24 | |
| *** fthiagogv__ has joined #openstack-manila | 18:26 | |
| *** porrua has quit IRC | 18:26 | |
| *** fthiagogv has quit IRC | 18:27 | |
| *** fthiagogv_ has quit IRC | 18:29 | |
| *** esker has joined #openstack-manila | 18:29 | |
| dencaval | bswartz, I'm a little bit confused about LVM driver for manila, one blueprint for removing and other for adding. | 18:29 |
| dencaval | bswartz, https://blueprints.launchpad.net/manila/+spec/remove-lvm-driver and https://blueprints.launchpad.net/manila/+spec/lvm-driver | 18:29 |
| bswartz | dencaval: I can provide the history there | 18:29 |
| *** 16WAAAVR4 has quit IRC | 18:30 | |
| *** esker has quit IRC | 18:30 | |
| *** esker has joined #openstack-manila | 18:31 | |
| bswartz | When we started the Manila project around Havana/Icehouse, we implemented an LVM driver to match what cinder had. However, the LVM driver didn't fit our original vision for how Manila should work -- we thought that all drivers should support share server, which the LVM driver could not, therefore we removed it during the Juno release cycle | 18:31 |
| bswartz | however in the Kilo release cycle, we realizes that it didn't make sense to require every driver to support share servers, so we introduce the share_servers=True/False modes | 18:32 |
| bswartz | we should have brought back the LVM driver at that time, but we didn't, because Generic was working well enough as a first-party driver | 18:32 |
| bswartz | During Liberty, the Generic driver started showing its limitations, so we've spend most of the Mitaka release adding more first party drivers | 18:33 |
| bswartz | we now have 4: Generic, LVM, LXD, and ZFS | 18:33 |
| bswartz | LVM is the simplest of the drivers | 18:33 |
| bswartz | and the most suitable for a new user who's just trying out Manila | 18:34 |
| bswartz | dencaval: hope that clears up any confusion | 18:34 |
| *** eharney has quit IRC | 18:34 | |
| dencaval | bswartz, Yes, It clears, thanks for sharing it. | 18:35 |
| dencaval | bswartz, Is there some table for pros/cons? | 18:35 |
| *** eharney has joined #openstack-manila | 18:37 | |
| *** bill_az has joined #openstack-manila | 18:41 | |
| bswartz | dencaval: I'm not sure it's written down somewhere | 18:46 |
| bswartz | the differences are many, and which differences are pros/cons depends on your goals | 18:46 |
| *** lgreg has left #openstack-manila | 18:49 | |
| dencaval | bswartz, I think I found the discussion: https://wiki.openstack.org/wiki/Manila/Networking | 18:53 |
| bswartz | dencaval: that's pretty old though | 18:54 |
| bswartz | dencaval: that's related to when we were deciding to to allow both share-server and no-share-server types of drivers | 18:54 |
| bswartz | dencaval: I will summarize the differences between the drivers | 18:55 |
| bswartz | Generic allows share-server and no-share-server modes, LVM is no-share-servers only | 18:55 |
| bswartz | Generic has dependencies on Nova, Glance, Neutron, and Cinder, LVM has no dependencies | 18:55 |
| bswartz | Generic allows you to take advantage of any storage controller that has a cinder driver, LVM just takes "local" storage and uses it to provide shares (although you can layer LVM on top of SAN storage) | 18:57 |
| bswartz | Generic is fairly complex and requires a lot of configuration, LVM is dead simple | 18:57 |
| bswartz | Both drivers are valuable and I hope they're both maintained for a long time | 18:58 |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Implement admin network in generic driver https://review.openstack.org/278699 | 18:58 |
| openstackgerrit | Goutham Pacha Ravi proposed openstack/manila: Scheduler enhancements for Share Replication https://review.openstack.org/277545 | 18:59 |
| dencaval | bswartz, Nice | 19:00 |
| *** lpetrut has joined #openstack-manila | 19:12 | |
| *** akerr has quit IRC | 19:14 | |
| *** akerr has joined #openstack-manila | 19:16 | |
| *** eharney has quit IRC | 19:16 | |
| openstackgerrit | Tiago Pasqualini da Silva proposed openstack/manila: Add separate process for Data Service copy https://review.openstack.org/279164 | 19:18 |
| *** jasonsb has quit IRC | 19:20 | |
| *** a_ta has joined #openstack-manila | 19:25 | |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: [Do not merge] Testing CI network LVM #1.4 https://review.openstack.org/281593 | 19:25 |
| *** cknight has joined #openstack-manila | 19:31 | |
| *** haomaiwang has joined #openstack-manila | 19:32 | |
| *** eharney has joined #openstack-manila | 19:33 | |
| *** dencaval has quit IRC | 19:37 | |
| *** haomaiwang has quit IRC | 19:37 | |
| *** dencaval has joined #openstack-manila | 19:37 | |
| *** lpetrut has quit IRC | 19:42 | |
| *** lpetrut has joined #openstack-manila | 19:46 | |
| *** ociuhandu has quit IRC | 19:58 | |
| *** zhonghua-lee has joined #openstack-manila | 20:07 | |
| *** bill_az has quit IRC | 20:12 | |
| *** zhonghua has quit IRC | 20:12 | |
| openstackgerrit | Valeriy Ponomaryov proposed openstack/manila: Add ZFSonLinux share driver https://review.openstack.org/277192 | 20:13 |
| *** sghatty has quit IRC | 20:15 | |
| openstackgerrit | Valeriy Ponomaryov proposed openstack/manila: Add ZFSonLinux share driver https://review.openstack.org/277192 | 20:16 |
| *** sghatty has joined #openstack-manila | 20:18 | |
| openstackgerrit | Dirk Mueller proposed openstack/manila: Generic driver: Fix unexport handling for nfs-utils 1.3.3 https://review.openstack.org/282045 | 20:19 |
| *** dencaval has quit IRC | 20:28 | |
| *** ociuhandu has joined #openstack-manila | 20:31 | |
| *** lgreg has joined #openstack-manila | 20:34 | |
| openstackgerrit | xing-yang proposed openstack/manila: Manage and unmanage snapshot https://review.openstack.org/249542 | 20:42 |
| *** lgreg has left #openstack-manila | 20:46 | |
| openstackgerrit | xing-yang proposed openstack/manila: Manage and unmanage snapshot https://review.openstack.org/249542 | 20:50 |
| xyang1 | bswartz: cknight vponomaryov, hi, I updated the patch: https://review.openstack.org/#/c/249542/ | 20:53 |
| xyang1 | bswartz: cknight vponomaryov moved provider_location to share_snapshot_instances table | 20:54 |
| *** porrua has joined #openstack-manila | 20:57 | |
| cknight | xyang1: thanks, I'll take a look | 20:57 |
| cknight | xyang1: but probably not before the 7 PM deadline! you're already good for the snapshot patch. | 20:58 |
| xyang1 | cknight: thanks | 20:59 |
| *** esker has quit IRC | 21:13 | |
| openstackgerrit | Andrei Ta proposed openstack/manila: Add additional documentation on extra spec operations. https://review.openstack.org/282063 | 21:14 |
| *** merooney has quit IRC | 21:15 | |
| *** martyturner has quit IRC | 21:29 | |
| *** martyturner has joined #openstack-manila | 21:30 | |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: [Do not merge] Testing CI Scenario tests https://review.openstack.org/281993 | 21:36 |
| *** timcl has quit IRC | 21:37 | |
| *** akerr has quit IRC | 21:38 | |
| *** ociuhandu has quit IRC | 21:41 | |
| openstackgerrit | Csaba Henk proposed openstack/manila: glusterfs.common: move the numreduct function to toplevel https://review.openstack.org/282069 | 21:43 |
| *** pgbridge has quit IRC | 21:44 | |
| openstackgerrit | Csaba Henk proposed openstack/manila: glusterfs: heketi: Add support for JWT Auth https://review.openstack.org/280486 | 21:44 |
| openstackgerrit | Csaba Henk proposed openstack/manila: glusterfs: introducing Heketi layout https://review.openstack.org/279090 | 21:44 |
| openstackgerrit | Clinton Knight proposed openstack/manila: Add preferred flag to export location metadata https://review.openstack.org/282070 | 21:53 |
| openstackgerrit | Ben Swartzlander proposed openstack/manila: Add doc for export location metadata https://review.openstack.org/282071 | 21:53 |
| *** merooney has joined #openstack-manila | 21:54 | |
| *** chlong has quit IRC | 21:57 | |
| *** chlong_ has joined #openstack-manila | 21:58 | |
| *** martyturner has quit IRC | 21:58 | |
| *** martyturner has joined #openstack-manila | 21:59 | |
| *** cknight has quit IRC | 21:59 | |
| *** pgbridge has joined #openstack-manila | 22:00 | |
| *** martyturner has quit IRC | 22:03 | |
| *** martyturner has joined #openstack-manila | 22:07 | |
| *** dustins has quit IRC | 22:09 | |
| *** ganso has quit IRC | 22:09 | |
| *** merooeny_ has joined #openstack-manila | 22:11 | |
| *** merooney has quit IRC | 22:14 | |
| *** eharney has quit IRC | 22:16 | |
| *** merooney has joined #openstack-manila | 22:20 | |
| *** porrua has quit IRC | 22:22 | |
| *** merooeny_ has quit IRC | 22:22 | |
| *** JoseMello has quit IRC | 22:28 | |
| openstackgerrit | Goutham Pacha Ravi proposed openstack/manila: Add Replication admin APIs and driver i/f changes https://review.openstack.org/282081 | 22:31 |
| *** cknight has joined #openstack-manila | 22:34 | |
| *** tpsilva has quit IRC | 22:36 | |
| *** ociuhandu has joined #openstack-manila | 22:41 | |
| *** cknight1 has joined #openstack-manila | 22:42 | |
| *** cknight has quit IRC | 22:44 | |
| *** lpetrut has quit IRC | 22:48 | |
| *** martyturner has quit IRC | 22:55 | |
| *** cknight1 has quit IRC | 23:01 | |
| *** xyang1 has quit IRC | 23:01 | |
| *** cknight has joined #openstack-manila | 23:11 | |
| *** lpabon has quit IRC | 23:11 | |
| *** gouthamr has quit IRC | 23:14 | |
| openstackgerrit | Goutham Pacha Ravi proposed openstack/manila: NetApp: Add Replication support in cDOT https://review.openstack.org/275416 | 23:16 |
| *** merooney has quit IRC | 23:20 | |
| *** cknight has quit IRC | 23:22 | |
| *** a_ta has quit IRC | 23:22 | |
| *** a_ta has joined #openstack-manila | 23:23 | |
| *** a_ta has quit IRC | 23:27 | |
| *** tobe has joined #openstack-manila | 23:28 | |
| openstackgerrit | Dirk Mueller proposed openstack/manila: Allow access to all test vm IPs in scenario testing https://review.openstack.org/282096 | 23:28 |
| *** ganso has joined #openstack-manila | 23:31 | |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila: [Do not merge] Testing CI network LVM #1.4 https://review.openstack.org/281593 | 23:31 |
| *** tobe has quit IRC | 23:35 | |
| *** dcain has quit IRC | 23:40 | |
| openstackgerrit | Csaba Henk proposed openstack/manila: glusterfs-native: use Heketi layout via self-signing https://review.openstack.org/282101 | 23:46 |
| *** openstackgerrit has quit IRC | 23:47 | |
| *** openstackgerrit_ is now known as openstackgerrit | 23:47 | |
| *** openstackgerrit_ has joined #openstack-manila | 23:48 | |
| *** openstackgerrit_ is now known as openstackgerrit | 23:48 | |
| *** openstackgerrit_ has joined #openstack-manila | 23:49 | |
| *** tobe has joined #openstack-manila | 23:49 | |
| openstackgerrit | Goutham Pacha Ravi proposed openstack/manila: Add tempest tests for Share Replication https://review.openstack.org/244843 | 23:52 |
| *** chlong_ has quit IRC | 23:52 | |
| *** openstackgerrit_ has quit IRC | 23:55 | |
| openstackgerrit | Csaba Henk proposed openstack/manila: glusterfs.common: move the numreduct function to toplevel https://review.openstack.org/282069 | 23:55 |
| openstackgerrit | Csaba Henk proposed openstack/manila: glusterfs_native: relocate module under glusterfs https://review.openstack.org/280935 | 23:55 |
| openstackgerrit | Csaba Henk proposed openstack/manila: glusterfs-native: use Heketi layout via self-signing https://review.openstack.org/282101 | 23:56 |
| openstackgerrit | Csaba Henk proposed openstack/manila: glusterfs: heketi: Add support for JWT Auth https://review.openstack.org/280486 | 23:56 |
| openstackgerrit | Csaba Henk proposed openstack/manila: glusterfs: introducing Heketi layout https://review.openstack.org/279090 | 23:56 |
| *** openstackgerrit_ has joined #openstack-manila | 23:57 | |
| *** sgotliv has quit IRC | 23:57 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!