*** haomaiwang has quit IRC | 00:01 | |
*** haomaiwang has joined #openstack-manila | 00:01 | |
*** martyturner has quit IRC | 00:01 | |
*** sedwards has joined #openstack-manila | 00:06 | |
openstackgerrit | Shaun Edwards proposed openstack/manila: EMC Isilon Driver Support For Read-Only Share https://review.openstack.org/271095 | 00:09 |
---|---|---|
*** darrenc_afk is now known as darrenc | 00:12 | |
*** sombrafam has quit IRC | 00:19 | |
*** sombrafam has joined #openstack-manila | 00:20 | |
*** merooney has quit IRC | 00:24 | |
*** jasonsb has joined #openstack-manila | 00:25 | |
*** sombrafam has quit IRC | 00:30 | |
*** sombrafam has joined #openstack-manila | 00:31 | |
*** lpabon has quit IRC | 00:44 | |
*** winston-1 has quit IRC | 00:58 | |
*** akerr has joined #openstack-manila | 00:59 | |
*** haomaiwang has quit IRC | 01:01 | |
*** haomaiwa_ has joined #openstack-manila | 01:01 | |
*** akerr_ has joined #openstack-manila | 01:02 | |
*** akerr has quit IRC | 01:05 | |
*** winston-d has joined #openstack-manila | 01:06 | |
*** akerr has joined #openstack-manila | 01:06 | |
*** akerr__ has joined #openstack-manila | 01:08 | |
*** akerr_ has quit IRC | 01:09 | |
*** akerr has quit IRC | 01:11 | |
*** dcain has joined #openstack-manila | 01:16 | |
*** haomaiwa_ has quit IRC | 01:17 | |
*** dcain1 has joined #openstack-manila | 01:26 | |
*** dcain has quit IRC | 01:29 | |
*** haomaiwang has joined #openstack-manila | 02:05 | |
*** mtanino has quit IRC | 02:07 | |
*** tingwang has joined #openstack-manila | 02:12 | |
openstackgerrit | ting wang proposed openstack/manila: Remove run_tests.sh wrapper https://review.openstack.org/270023 | 02:22 |
*** sombrafam has quit IRC | 02:41 | |
*** sombrafam has joined #openstack-manila | 02:42 | |
*** tingwang has quit IRC | 02:43 | |
*** sombrafam has quit IRC | 02:47 | |
*** sombrafam has joined #openstack-manila | 02:48 | |
*** vbellur has joined #openstack-manila | 02:53 | |
*** haomaiwang has quit IRC | 03:01 | |
*** tingwang has joined #openstack-manila | 03:01 | |
*** 77CAAARUP has joined #openstack-manila | 03:01 | |
openstackgerrit | ting wang proposed openstack/manila: isoformat instead of deprecated timeutils.isotime https://review.openstack.org/271125 | 03:13 |
*** cknight has joined #openstack-manila | 03:14 | |
*** tingwang has quit IRC | 03:14 | |
openstackgerrit | liucheng proposed openstack/manila: Huawei: Create share from snapshot support in Huawei driver https://review.openstack.org/259917 | 03:27 |
*** bill_az has quit IRC | 03:46 | |
*** cknight has quit IRC | 03:59 | |
*** 77CAAARUP has quit IRC | 04:01 | |
*** haomaiwa_ has joined #openstack-manila | 04:01 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/manila: Updated from global requirements https://review.openstack.org/270944 | 04:04 |
*** akshai has joined #openstack-manila | 04:13 | |
*** rraja|afk is now known as rraja | 04:16 | |
*** gouthamr has joined #openstack-manila | 04:18 | |
*** akshai has quit IRC | 04:18 | |
*** gouthamr_ has joined #openstack-manila | 04:19 | |
*** gouthamr has quit IRC | 04:22 | |
*** vbellur has quit IRC | 04:52 | |
*** haomaiwa_ has quit IRC | 05:01 | |
*** esker has joined #openstack-manila | 05:01 | |
*** haomaiwang has joined #openstack-manila | 05:01 | |
*** haomaiwang has quit IRC | 05:13 | |
*** akshai has joined #openstack-manila | 05:14 | |
*** akshai has quit IRC | 05:19 | |
*** sgotliv has joined #openstack-manila | 05:23 | |
*** esker has quit IRC | 05:32 | |
*** gouthamr_ has quit IRC | 05:40 | |
openstackgerrit | xing-yang proposed openstack/manila: Manage and unmanage snapshot https://review.openstack.org/249542 | 05:50 |
openstackgerrit | Ramana Raja proposed openstack/manila: Enable and add Tempest tests for cephfs protocol https://review.openstack.org/270713 | 06:01 |
*** akshai has joined #openstack-manila | 06:15 | |
*** akshai has quit IRC | 06:20 | |
*** akerr__ is now known as akerr_away | 06:38 | |
*** nkrinner has joined #openstack-manila | 06:38 | |
*** tingwang has joined #openstack-manila | 07:04 | |
*** lpetrut has joined #openstack-manila | 07:16 | |
*** akshai has joined #openstack-manila | 07:16 | |
*** akshai has quit IRC | 07:21 | |
*** sombrafam has quit IRC | 07:39 | |
*** sombrafam has joined #openstack-manila | 07:40 | |
*** houming has joined #openstack-manila | 07:42 | |
*** ociuhandu has joined #openstack-manila | 07:43 | |
*** sombrafam has quit IRC | 07:48 | |
*** sombrafam has joined #openstack-manila | 07:48 | |
openstackgerrit | liucheng proposed openstack/manila: Huawei: Create share from snapshot support in Huawei driver https://review.openstack.org/259917 | 07:48 |
*** lpetrut has quit IRC | 07:55 | |
openstackgerrit | zhaohua proposed openstack/manila: Change the condition of using max_over_subscription_ratio https://review.openstack.org/271172 | 07:58 |
*** akshai has joined #openstack-manila | 08:17 | |
*** akshai has quit IRC | 08:22 | |
*** baojg has joined #openstack-manila | 08:44 | |
*** akshai has joined #openstack-manila | 09:18 | |
openstackgerrit | zhaohua proposed openstack/manila: Change the condition of using max_over_subscription_ratio https://review.openstack.org/271172 | 09:20 |
*** akshai has quit IRC | 09:23 | |
kaisers1 | vponomaryov: ping | 09:23 |
openstackgerrit | Ramana Raja proposed openstack/manila: Add Ceph Native driver https://review.openstack.org/270211 | 09:25 |
openstackgerrit | Ramana Raja proposed openstack/manila: Enable and add Tempest tests for cephfs protocol https://review.openstack.org/270713 | 09:25 |
*** tpsilva has joined #openstack-manila | 09:44 | |
*** baojg has quit IRC | 09:52 | |
*** baojg has joined #openstack-manila | 09:54 | |
openstackgerrit | ting wang proposed openstack/manila: TrivialFix: Remove deprecated option 'DEFAULT/verbose' https://review.openstack.org/270069 | 09:56 |
vponomaryov | kaisers1: pong | 09:58 |
*** haomaiwang has joined #openstack-manila | 09:59 | |
*** haomaiwang has quit IRC | 10:01 | |
*** haomaiwa_ has joined #openstack-manila | 10:01 | |
openstackgerrit | Tiago Pasqualini da Silva proposed openstack/manila: Hitachi HNAS driver share shrink https://review.openstack.org/265434 | 10:01 |
openstackgerrit | ting wang proposed openstack/manila: TrivialFix: Remove deprecated option 'DEFAULT/verbose' https://review.openstack.org/270069 | 10:02 |
*** akshai has joined #openstack-manila | 10:18 | |
*** akshai has quit IRC | 10:23 | |
*** jasonsb has quit IRC | 10:26 | |
openstackgerrit | Valeriy Ponomaryov proposed openstack/manila: Add admin network for DHSS=True share drivers https://review.openstack.org/260021 | 10:30 |
*** haomaiwa_ has quit IRC | 10:32 | |
*** esker has joined #openstack-manila | 10:33 | |
*** sombrafam has quit IRC | 10:34 | |
*** sombrafam has joined #openstack-manila | 10:35 | |
*** sgotliv has quit IRC | 10:37 | |
openstackgerrit | Merged openstack/manila-ui: Updated from global requirements https://review.openstack.org/268459 | 10:38 |
*** lpetrut has joined #openstack-manila | 10:38 | |
*** esker has quit IRC | 10:38 | |
*** lpetrut has quit IRC | 10:40 | |
*** lpetrut has joined #openstack-manila | 10:40 | |
*** sombrafam has quit IRC | 10:40 | |
*** sombrafam has joined #openstack-manila | 10:42 | |
*** tingwang has quit IRC | 10:47 | |
*** baojg has quit IRC | 10:58 | |
kaisers1 | vponomaryov: sorry, had a major issue popping up in our office. I'm all here now. | 11:14 |
kaisers1 | vponomaryov: q was if you have a minute to spend on some questions & thoughts on Share Servers | 11:15 |
vponomaryov | kaisers1: minute will be enough? )) | 11:15 |
kaisers1 | vponomaryov: :) | 11:18 |
kaisers1 | vponomaryov: probably not? dunno, let's see | 11:18 |
vponomaryov | you owe me 2 minutes already ) | 11:19 |
kaisers1 | vponomaryov: So, share servers are there to provide network integration and for some storages the NFS server. Are there other reasons? | 11:19 |
openstackgerrit | Valeriy Ponomaryov proposed openstack/manila-image-elements: [DEBUG] do not merge https://review.openstack.org/271246 | 11:19 |
vponomaryov | kaisers1: you mean reason to have share servers? | 11:20 |
kaisers1 | We've been discussing if there's a way to omit share servers and provide access to the backend by the driver e.g. only configuring the required neutron networks | 11:20 |
kaisers1 | vponomaryov: yep | 11:20 |
vponomaryov | kaisers1: to have resource separation by network and computing | 11:20 |
vponomaryov | kaisers1: for different tenants | 11:21 |
*** sombrafam has quit IRC | 11:21 | |
kaisers1 | vponomaryov: right. That would've to be to be done in the networks | 11:21 |
kaisers1 | e.g. one network per tenant | 11:21 |
vponomaryov | kaisers1: why one network per tenant? | 11:21 |
vponomaryov | kaisers1: allow to use any networks - that is the idea | 11:22 |
kaisers1 | vponomaryov: to ensure separation of different tenants? | 11:22 |
vponomaryov | kaisers1: user may want to use one common net and one private | 11:22 |
vponomaryov | kaisers1: for different shares | 11:22 |
kaisers1 | vponomaryov: hmm, I'll draw up some usage scenarios regarding that, right | 11:23 |
kaisers1 | vponomaryov: the basic idea is that we could run without a share server if we can get the neutron / network configs right | 11:23 |
vponomaryov | kaisers1: you already can | 11:24 |
vponomaryov | kaisers1: using DHSS=False mode | 11:24 |
vponomaryov | kaisers1: what is the problem? | 11:24 |
kaisers1 | vponomaryov: ok, i was not clear about that. it's one of the questions. The docs have different comments on wether dhss = false uses no or a generic share server | 11:24 |
*** ociuhandu has quit IRC | 11:25 | |
vponomaryov | kaisers1: "on wether dhss = false uses no or a generic share server"? | 11:25 |
vponomaryov | kaisers1: you mean it is not clear for you what is the diff between DHSS=False and DHSS=True driver modes? | 11:26 |
*** fthiagogv has joined #openstack-manila | 11:26 | |
kaisers1 | vponomaryov: regarding that point the question came up, yep *duck* | 11:26 |
vponomaryov | kaisers1: DHSS = "driver handles share servers" - here is diff | 11:27 |
kaisers1 | I know that DHSS=true requires the driver to manage the share servers but .. | 11:27 |
vponomaryov | kaisers1: so, everything up to out-of-manila env for your storage | 11:27 |
vponomaryov | kaisers1: but what? | 11:29 |
kaisers1 | vponomaryov: you mean with DHSS=True means manila does not care about the Share Servers but expects the driver to start and manage them | 11:29 |
kaisers1 | vponomaryov: sorry, stopped the sentence at but because i was reading & thinking about your comment | 11:29 |
vponomaryov | kaisers1: I mean with DHSS=True means manila DOES CARE about the Share Servers | 11:30 |
vponomaryov | kaisers1: in DHSS=False does not, so, driver makes less actions, simpler logic, common things for different tenants | 11:31 |
kaisers1 | vponomaryov: ok and with DHSS=Flase there's no (generic or whatever) Share Server but only the network config that has to be brought up to access the storage backend | 11:31 |
kaisers1 | s/Flase/False/ | 11:31 |
vponomaryov | kaisers1: yes, we can say so | 11:31 |
vponomaryov | kaisers1: JFYI, "share server" is definition of manila to name thing, that it created and going to track its lifecycle | 11:32 |
vponomaryov | kaisers1: some "host" you have always | 11:32 |
kaisers1 | vponomaryov: I understood it's a VM in the generic manila or tenant network | 11:33 |
vponomaryov | kaisers1: just in some cases it is handled by manila in other - not | 11:33 |
vponomaryov | kaisers1: generic manila? | 11:33 |
vponomaryov | kaisers1: there is the only Manila! | 11:33 |
vponomaryov | kaisers1: )) | 11:34 |
kaisers1 | vponomaryov: This generic manila share server seems to be something i got totally wrong some months ago, sorry | 11:34 |
vponomaryov | kaisers1: I guess you are talking about "generic driver" for manila? | 11:34 |
vponomaryov | kaisers1: this driver tied on Nova and Cinder | 11:34 |
kaisers1 | vponomaryov: died? | 11:34 |
kaisers1 | vponomaryov: ties together, oh, ok | 11:35 |
*** sombrafam has joined #openstack-manila | 11:35 | |
vponomaryov | kaisers1: so, it can use only Nova VMs, nothing else | 11:35 |
vponomaryov | kaisers1: LVM driver uses manila-share host in DHSS=False driver | 11:36 |
*** houming has quit IRC | 11:36 | |
vponomaryov | kaisers1: BUT | 11:36 |
vponomaryov | kaisers1: noone restricts admin to configure it inside of Nova VM | 11:36 |
vponomaryov | kaisers1: in that case Nova VM just should be powerful enough to handle load | 11:37 |
kaisers1 | vponomaryov: Does this mean a Share Server might be e.g. a baremetal server outside of OpenStack? | 11:38 |
vponomaryov | kaisers1: can be baremetal, but, I guess, it is incorrect to say "outside of OpenStack". | 11:39 |
vponomaryov | kaisers1: it is env for service | 11:39 |
vponomaryov | kaisers1: it cannot be outside | 11:39 |
vponomaryov | kaisers1: distributed - yes | 11:39 |
vponomaryov | kaisers1: in several regions | 11:39 |
kaisers1 | vponomaryov: thinking about this.... | 11:40 |
vponomaryov | kaisers1: outside of OpenStack we mean - out of OpenStack admin control | 11:40 |
kaisers1 | vponomaryov: ok, to be more specific. A Share Server has to be under the OpenStack admin control and it has to be reachable via Neutron network. | 11:41 |
*** nidhimittalhada has joined #openstack-manila | 11:41 | |
kaisers1 | vponomaryov: For the latter there are more details to be clarified but as a basic statement, would that be more correct^? | 11:41 |
vponomaryov | kaisers1: no | 11:41 |
vponomaryov | kaisers1: we have "network plugin" approach for share servers, and it can be anything, not only Neutron | 11:42 |
vponomaryov | kaisers1: any Manila resources as all other projects resources shoudl be available for admin influence | 11:43 |
kaisers1 | vponomaryov: yes. It's just that i'm focussing on Neutron setups in my thoughts currenlty | 11:43 |
kaisers1 | vponomaryov: so a driver using DHSS=True and using e.g. 1 to n static Share Servers that are specified to the driver via config file is such a scenario? | 11:45 |
kaisers1 | vponomaryov: in this example static Share Servers beeing bare metal boxes reachable via neutron networks for the VMs... | 11:46 |
*** openstackgerrit has quit IRC | 11:47 | |
*** openstackgerrit has joined #openstack-manila | 11:48 | |
*** timcl has joined #openstack-manila | 11:48 | |
vponomaryov | kaisers1: again, "share server" is definition of manila for naming compute/networking things that manila creates and handles | 11:51 |
vponomaryov | kaisers1: they are not specified via config | 11:51 |
kaisers1 | vponomaryov: ok, that was wrong | 11:52 |
*** timcl1 has joined #openstack-manila | 11:52 | |
*** timcl has quit IRC | 11:52 | |
kaisers1 | Thanks for your answers so far and (rather rethorical) 'minute'. I'll re-read some stuff with this input. | 11:53 |
*** haomaiwang has joined #openstack-manila | 11:54 | |
*** haomaiwang has quit IRC | 11:58 | |
*** ociuhandu has joined #openstack-manila | 12:02 | |
*** erlon has joined #openstack-manila | 12:10 | |
*** JoseMello has joined #openstack-manila | 12:11 | |
*** akshai has joined #openstack-manila | 12:20 | |
*** akshai has quit IRC | 12:24 | |
nidhimittalhada | @vponomaryov: Thanks. | 12:25 |
*** nidhimittalhada has left #openstack-manila | 12:25 | |
*** openstackgerrit has quit IRC | 12:33 | |
*** openstackgerrit has joined #openstack-manila | 12:33 | |
*** haomaiwang has joined #openstack-manila | 12:55 | |
*** bill_az has joined #openstack-manila | 12:59 | |
*** haomaiwang has quit IRC | 13:00 | |
*** timcl1 has quit IRC | 13:12 | |
openstackgerrit | Valeriy Ponomaryov proposed openstack/manila-image-elements: [DEBUG] do not merge https://review.openstack.org/271246 | 13:14 |
*** esker has joined #openstack-manila | 13:15 | |
*** vbellur has joined #openstack-manila | 13:15 | |
openstackgerrit | Valeriy Ponomaryov proposed openstack/manila-image-elements: [DEBUG2] do not merge https://review.openstack.org/271281 | 13:19 |
*** esker has quit IRC | 13:20 | |
*** timcl has joined #openstack-manila | 13:32 | |
*** akerr_away is now known as akerr__ | 13:37 | |
openstackgerrit | Valeriy Ponomaryov proposed openstack/manila-image-elements: [DEBUG3] do not merge https://review.openstack.org/271288 | 13:38 |
openstackgerrit | Valeriy Ponomaryov proposed openstack/manila-image-elements: [DEBUG4] do not merge https://review.openstack.org/271290 | 13:39 |
*** akerr__ is now known as akerr | 13:39 | |
openstackgerrit | Valeriy Ponomaryov proposed openstack/manila-image-elements: [DEGUG5] do not merge https://review.openstack.org/271297 | 13:47 |
openstackgerrit | Csaba Henk proposed openstack/manila: gluster*: add proper getter/setters for volume options https://review.openstack.org/271303 | 13:52 |
*** esker has joined #openstack-manila | 13:56 | |
*** sombrafam has quit IRC | 13:58 | |
*** nkrinner has quit IRC | 14:01 | |
*** lpabon has joined #openstack-manila | 14:02 | |
*** esker has quit IRC | 14:03 | |
openstackgerrit | Valeriy Ponomaryov proposed openstack/manila-image-elements: Fix buildimage job https://review.openstack.org/271309 | 14:04 |
*** esker has joined #openstack-manila | 14:06 | |
openstackgerrit | Valeriy Ponomaryov proposed openstack/manila-image-elements: Update system hosts file to understand local domainnames https://review.openstack.org/269734 | 14:07 |
openstackgerrit | Valeriy Ponomaryov proposed openstack/manila-image-elements: Fix launchpad url in CONTRIBUTING.rst file https://review.openstack.org/270805 | 14:07 |
openstackgerrit | Valeriy Ponomaryov proposed openstack/manila-image-elements: Set bash as default shell https://review.openstack.org/269753 | 14:07 |
openstackgerrit | Valeriy Ponomaryov proposed openstack/manila-image-elements: Install vim text editor https://review.openstack.org/269766 | 14:07 |
*** absubram has joined #openstack-manila | 14:10 | |
*** absubram_ has joined #openstack-manila | 14:11 | |
*** haomaiwang has joined #openstack-manila | 14:14 | |
*** absubram has quit IRC | 14:15 | |
*** absubram_ is now known as absubram | 14:15 | |
*** porrua has joined #openstack-manila | 14:17 | |
*** haomaiwang has quit IRC | 14:18 | |
*** akshai has joined #openstack-manila | 14:21 | |
*** porrua has quit IRC | 14:22 | |
*** absubram has quit IRC | 14:23 | |
*** gouthamr has joined #openstack-manila | 14:24 | |
*** absubram has joined #openstack-manila | 14:24 | |
*** gouthamr_ has joined #openstack-manila | 14:25 | |
*** porrua has joined #openstack-manila | 14:25 | |
*** kaisers1 has quit IRC | 14:27 | |
*** kaisers has joined #openstack-manila | 14:27 | |
*** gouthamr has quit IRC | 14:28 | |
*** esker has quit IRC | 14:31 | |
*** akshai has quit IRC | 14:31 | |
*** cknight has joined #openstack-manila | 14:33 | |
*** haomaiwang has joined #openstack-manila | 14:35 | |
*** xyang1 has joined #openstack-manila | 14:35 | |
*** dustins has joined #openstack-manila | 14:35 | |
*** haomaiwang has quit IRC | 14:39 | |
*** xyang_ has joined #openstack-manila | 14:41 | |
*** pcaruana has joined #openstack-manila | 14:46 | |
openstackgerrit | Emilien Macchi proposed openstack/puppet-manila: drop mock dependency https://review.openstack.org/271338 | 14:50 |
*** haomaiwang has joined #openstack-manila | 14:56 | |
bswartz | vponomaryov: I see a job that ran on the node "bare-trusty-hpcloud-b3-7398271" | 14:57 |
bswartz | that implies that hpcloud is still being used for openstack resources | 14:57 |
vponomaryov | bswartz: what's wrong with it? | 14:58 |
bswartz | hp announced that they were shutting their cloud down | 14:58 |
bswartz | I thought it was gone already but apparently not | 14:58 |
*** haomaiwang has quit IRC | 15:00 | |
*** timcl has quit IRC | 15:03 | |
*** esker has joined #openstack-manila | 15:04 | |
*** timcl1 has joined #openstack-manila | 15:08 | |
*** cknight has quit IRC | 15:10 | |
*** cknight has joined #openstack-manila | 15:13 | |
timcl1 | bswartz: I think the HP cloud is around to the 29th | 15:15 |
openstackgerrit | Valeriy Ponomaryov proposed openstack/manila-image-elements: [DEGUG5] do not merge https://review.openstack.org/271297 | 15:16 |
*** haomaiwang has joined #openstack-manila | 15:17 | |
bswartz | timcl1: oh so maybe the zuul resources for openstack will significantly decrease at the end of the month? | 15:22 |
*** haomaiwang has quit IRC | 15:22 | |
openstackgerrit | Valeriy Ponomaryov proposed openstack/manila-image-elements: [DEBUG6] do not merge https://review.openstack.org/271375 | 15:26 |
*** rraja has quit IRC | 15:30 | |
*** haomaiwang has joined #openstack-manila | 15:38 | |
*** absubram has quit IRC | 15:40 | |
*** haomaiwang has quit IRC | 15:43 | |
*** haomaiwang has joined #openstack-manila | 15:58 | |
*** timcl1 has quit IRC | 15:59 | |
*** sombrafam has joined #openstack-manila | 15:59 | |
*** haomaiwang has quit IRC | 16:01 | |
*** timcl has joined #openstack-manila | 16:06 | |
*** cknight1 has joined #openstack-manila | 16:11 | |
*** cknight has quit IRC | 16:14 | |
*** ociuhandu has quit IRC | 16:19 | |
*** a_ta_ has joined #openstack-manila | 16:24 | |
*** a_ta has quit IRC | 16:27 | |
openstackgerrit | xing-yang proposed openstack/manila: Tempest tests for Manage and unmanage snapshot https://review.openstack.org/262114 | 16:35 |
*** mtanino has joined #openstack-manila | 16:43 | |
*** timcl has quit IRC | 16:45 | |
*** esker has quit IRC | 16:59 | |
*** absubram has joined #openstack-manila | 17:10 | |
openstackgerrit | Merged openstack/manila: Hide snapshots with no instances from listing https://review.openstack.org/261289 | 17:10 |
*** timcl has joined #openstack-manila | 17:21 | |
*** haomaiwa_ has joined #openstack-manila | 17:21 | |
*** a_ta_ has quit IRC | 17:24 | |
*** a_ta has joined #openstack-manila | 17:24 | |
*** timcl has left #openstack-manila | 17:25 | |
*** haomaiwa_ has quit IRC | 17:26 | |
*** esker has joined #openstack-manila | 17:27 | |
*** a_ta_ has joined #openstack-manila | 17:27 | |
*** timcl1 has joined #openstack-manila | 17:27 | |
*** absubram has quit IRC | 17:29 | |
*** a_ta has quit IRC | 17:29 | |
*** fthiagogv has quit IRC | 17:33 | |
openstackgerrit | Shaun Edwards proposed openstack/manila: EMC Isilon Driver Support For Read-Only Share https://review.openstack.org/271095 | 17:33 |
*** sombrafam has quit IRC | 17:39 | |
*** jasonsb has joined #openstack-manila | 17:40 | |
*** haomaiwang has joined #openstack-manila | 17:42 | |
*** esker has quit IRC | 17:44 | |
*** haomaiwang has quit IRC | 17:47 | |
-openstackstatus- NOTICE: Restarting zuul due to a memory leak | 17:50 | |
*** esker has joined #openstack-manila | 17:52 | |
*** sombrafam has joined #openstack-manila | 18:00 | |
*** a_ta_ has quit IRC | 18:02 | |
*** a_ta has joined #openstack-manila | 18:03 | |
*** a_ta has joined #openstack-manila | 18:03 | |
*** 14WAAR75L has joined #openstack-manila | 18:03 | |
*** esker has quit IRC | 18:05 | |
*** 14WAAR75L has quit IRC | 18:08 | |
*** absubram has joined #openstack-manila | 18:11 | |
*** martyturner has joined #openstack-manila | 18:17 | |
*** haomaiwang has joined #openstack-manila | 18:24 | |
*** porrua has quit IRC | 18:27 | |
*** haomaiwang has quit IRC | 18:29 | |
*** lpabon has quit IRC | 18:40 | |
*** haomaiwang has joined #openstack-manila | 18:45 | |
*** martyturner has quit IRC | 18:49 | |
*** haomaiwang has quit IRC | 18:49 | |
*** sombrafam has quit IRC | 18:51 | |
*** sombrafam has joined #openstack-manila | 18:53 | |
*** dcain has joined #openstack-manila | 18:57 | |
*** sombrafam has quit IRC | 18:59 | |
*** dcain1 has quit IRC | 19:00 | |
*** sombrafam has joined #openstack-manila | 19:00 | |
*** lpetrut has quit IRC | 19:04 | |
bswartz | vponomaryov: I've installed ubuntu 14.04 and 16.04 so far | 19:05 |
bswartz | 16.04 has all the right zfs packages, 14.04 definitely does not | 19:05 |
bswartz | I'm going to take a look at 15.10 next | 19:05 |
*** haomaiwang has joined #openstack-manila | 19:06 | |
*** haomaiwang has quit IRC | 19:11 | |
*** esker has joined #openstack-manila | 19:13 | |
openstackgerrit | Goutham Pacha Ravi proposed openstack/manila: Fix db shim layer mismatches with implementation https://review.openstack.org/269681 | 19:13 |
*** xyang_ has quit IRC | 19:17 | |
*** dcain1 has joined #openstack-manila | 19:19 | |
*** dcain has quit IRC | 19:23 | |
*** haomaiwang has joined #openstack-manila | 19:27 | |
*** jasonsb has quit IRC | 19:27 | |
*** lpetrut has joined #openstack-manila | 19:28 | |
bswartz | vponomaryov: 15.10 has the packages | 19:29 |
bswartz | I don't think there's any point to looking at 14.10 or 15.04 | 19:29 |
bswartz | the only reasonable choices for testing at 15.10 or 16.04 (dev builds) | 19:30 |
openstackgerrit | Tiago Pasqualini da Silva proposed openstack/manila: Add update_access() method to driver interface https://review.openstack.org/245126 | 19:30 |
tpsilva | bswartz: ^ would appreciate any reviews | 19:31 |
*** haomaiwang has quit IRC | 19:32 | |
bswartz | tpsilva: what's changed? | 19:34 |
*** lpetrut has quit IRC | 19:37 | |
tbarron | bswartz: I think I missed the motivation for manila having a ZFS dependency. | 19:40 |
tbarron | was this discussed at mid-cycle or meetings and I was snoozing? | 19:40 |
markstur | tbarron, When you need a nap, you need a nap | 19:41 |
tbarron | zzzzz fs | 19:41 |
tbarron | markstur: you probably know the answer ... | 19:42 |
markstur | markstur, I would have answered if I was sure. Didn't want to expose myself. I'm thinking it has something to do with 1st party drivers. | 19:43 |
*** xyang_ has joined #openstack-manila | 19:44 | |
tpsilva | bswartz: I implemented the compatbility with older api versions, as discussed on the weekly meeting | 19:45 |
*** martyturner has joined #openstack-manila | 19:46 | |
*** haomaiwang has joined #openstack-manila | 19:48 | |
*** haomaiwang has quit IRC | 19:53 | |
*** sombrafam has quit IRC | 19:55 | |
tbarron | markstur: yeah, my guess was a DHSS=true first-party alternative to the LXC driver | 19:57 |
tbarron | I'd like the first-party drivers to run on the full range of devstack supported platforms. | 19:58 |
bswartz | tbarron: replication in first party drivers | 19:59 |
bswartz | tbarron: the original plan had been drbd, but experiments show drbd to be a resource hog | 20:00 |
tbarron | ah, that makes sense. | 20:00 |
bswartz | so zfs is beleived to be a lighter-weight alternative that should serve our purposes | 20:00 |
bswartz | right now i'm struggling with poor zfs support on Linux in general | 20:01 |
tbarron | bswartz: thank Linus, or Oracle, depending on your theory, for that. | 20:01 |
bswartz | well I'm going to thank shuttleworth for finally making it NOT suck in the upcoming Ubuntu LTS release | 20:02 |
bswartz | the challenge we will face is what to do between now and the LTS release | 20:02 |
tbarron | there's kernel support for zfs in ubuntu? | 20:02 |
bswartz | tbarron: since 15.10 at least, maybe older | 20:02 |
bswartz | 16.04 works perfectly | 20:02 |
bswartz | I'm struggling with 15.10 | 20:03 |
tbarron | but presumably not in rhel or suse | 20:03 |
tbarron | not trying to be difficult here, just figuring things out. | 20:03 |
bswartz | http://paste.openstack.org/show/484739/ | 20:03 |
bswartz | I'm not familiar with the RHEL+ZFS situation but if I had to guess I would say they probably don't work well together | 20:04 |
bswartz | basing that guess on RedHat's tendency to avoid stuff that's not "pure" FOSS | 20:04 |
*** lpetrut has joined #openstack-manila | 20:05 | |
tbarron | well, you know my interest in not only having stuff that works for CI but also for POCs, troubleshooting, etc. | 20:05 |
bswartz | tbarron: you may need to debase yourself and use Ubuntu guest images to debug the Manila generic driver if this ZFS experiment turns out well | 20:06 |
bswartz | the happiest outcome would be for fedora to add ZFS support in the near future though | 20:06 |
bswartz | (if it's not already there) | 20:06 |
tbarron | bswartz: I will check. | 20:07 |
bswartz | F22 has zfs-fuse but no kernel support | 20:07 |
* bswartz runs F22 currently | 20:08 | |
tbarron | but I have to ask, have you considered btrfs for this job? | 20:08 |
bswartz | tbarron: nobody believes it has the maturity | 20:08 |
tbarron | for manila? just joking. | 20:08 |
*** haomaiwang has joined #openstack-manila | 20:09 | |
tbarron | f23 also is fuse | 20:09 |
bswartz | I mean it's worth considering, but btrfs doesn't appear to offer any advantages compared to zfs, and it's known to be less mature by 5+ years | 20:10 |
tbarron | well, the advantage is cross-platform, open source; the disadvantage, as you say, is likely still its maturity. | 20:12 |
tbarron | I wouldn't run a bank DB on it today for sure. | 20:13 |
*** haomaiwang has quit IRC | 20:13 | |
bswartz | zfs is WAY more cross platform than btrfs | 20:13 |
bswartz | zfs also runs on OpenSolaris, FreeBSD, etc -- I'm pretty sure btrfs is linux only | 20:14 |
bswartz | the license is one definite advantage | 20:14 |
*** lpabon has joined #openstack-manila | 20:14 | |
tbarron | ok, cross-distribution/platform with kernel support within linux | 20:14 |
*** pcaruana has quit IRC | 20:14 | |
tbarron | I *like* ZFS, that's not the point. | 20:15 |
bswartz | well at this point I'll take anything that works | 20:15 |
tbarron | And, as I say, not trying to be argumentative or obstructionist. Just recording a concern. | 20:15 |
bswartz | I want replication support merged in manila in the next 3 weeks at the latest | 20:16 |
*** martyturner has quit IRC | 20:16 | |
bswartz | so whatever we can get working in that time will have to do | 20:16 |
*** martyturner has joined #openstack-manila | 20:16 | |
bswartz | woohoo I think I solved my problem on 15.10 | 20:16 |
*** esker has quit IRC | 20:16 | |
bswartz | so we may not be forced to use ubuntu bleeding-edge | 20:17 |
*** esker has joined #openstack-manila | 20:17 | |
bswartz | tbarron: while rpms for zfs may not exist on fedora, I bet you can find recipes to download and build kernel modules for it that are not too hard to implement | 20:17 |
bswartz | the same case seems to exist for ubuntu versions going back as far as 12.04 | 20:18 |
tbarron | can user-space zfs work for POC on this? | 20:19 |
bswartz | tbarron: possibly.... | 20:19 |
tbarron | Not necessarily objecting to buildiing custom kernel, just exploring the possibility space. | 20:19 |
bswartz | I don't intend to test the fuse versions of zfs though since I have working as kernel modules in both 15.10 and 16.04' | 20:20 |
tbarron | As i say I'm looking for simple, easily-reproducible POC setup, not necessarily performance. | 20:20 |
tbarron | understood | 20:20 |
bswartz | I agree | 20:20 |
*** timcl1 has quit IRC | 20:20 | |
bswartz | the drbd experiment shows us that performance isn't something we can ignore though | 20:20 |
tbarron | so what happened? if you have time. | 20:21 |
bswartz | our gate jobs run using crappy VMs inside crappy vms | 20:21 |
bswartz | not enough RAM, not enough CPU to do much | 20:21 |
tbarron | sorry to interject, but isn't zfs a memory hog? | 20:21 |
bswartz | consider that zuul job runner nodes only have 4GB and we have to run multiple VMs inside that | 20:21 |
bswartz | tbarron: it can be | 20:22 |
bswartz | so part of this experiment is to see how much we can squeeze it down | 20:22 |
tbarron | kk, makes sense | 20:22 |
bswartz | tbarron: do you know anything about the memory characteristics of btrfs? I don't see any reason to expect it to be better | 20:25 |
*** timcl1 has joined #openstack-manila | 20:25 | |
*** jasonsb has joined #openstack-manila | 20:27 | |
tbarron | *know* would be a strong word. btrfs advocates claim it uses significantly less memory. | 20:27 |
tbarron | there's a lot of religion out there on this subject. | 20:27 |
tbarron | if I find something solid I'll share. | 20:27 |
bswartz | I'm reading about ZFS's RAM requirements and btrfs and if we run into RAM issues, then btrfs could get us out of jail | 20:28 |
bswartz | I just don't think I could recommend btrfs for production workloads with a straight face | 20:28 |
tbarron | I know you are looking at a three week implementation, but that leads naturally to a question of | 20:29 |
tbarron | what api you have to the underlying replication file system, how easy to slide one in and one out. | 20:29 |
*** haomaiwang has joined #openstack-manila | 20:30 | |
bswartz | tbarron: the goal isn't to have the driver done in 3 weeks, the goal is to have a POC that makes us feel okay about merging the core feature | 20:30 |
bswartz | we have until feature freeze to merge replication support in the driver | 20:30 |
*** martyturner has quit IRC | 20:32 | |
*** haomaiwang has quit IRC | 20:34 | |
*** lpetrut has quit IRC | 20:53 | |
*** mtanino has quit IRC | 20:55 | |
*** timcl1 has left #openstack-manila | 20:59 | |
openstackgerrit | Merged openstack/manila: Add LVM driver https://review.openstack.org/232970 | 21:01 |
*** akerr_ has joined #openstack-manila | 21:09 | |
*** akerr has quit IRC | 21:11 | |
markstur | gouthamr_, Did you figure out if you get credit now for Co-Author? I see it is fixed for Akshai | 21:13 |
bswartz | markstur: what kind of credit? | 21:14 |
markstur | bswartz, Stackalytics kind of credit | 21:14 |
gouthamr_ | markstur: i guess i messed up with a git reset at some point.. | 21:14 |
markstur | bswartz, Not sure how co-author works | 21:14 |
bswartz | stackalytics is one kind of credit, but what most people care about is ATC status with the foundation | 21:15 |
markstur | bswartz, gouthamr_ I'm guessing Goutham gets paid either way | 21:15 |
bswartz | yeah obviously | 21:15 |
gouthamr_ | markstur: lol | 21:15 |
*** jasonsb has quit IRC | 21:15 | |
gouthamr_ | markstur bswartz: though i still dunno how owner is different from author.. | 21:16 |
markstur | bswartz, Yes. Goutham probably doesn't care much about 1 more commit. Akshai might. Just nice to get it right -- and I don't have the co-author thing figured out | 21:16 |
bswartz | owner exists only in gerrit, not in git, and it's immutable | 21:16 |
*** marcusvrn_ has quit IRC | 21:17 | |
bswartz | basically the committer of any new change in gerrit becomes the owner, and later commits don't change that | 21:17 |
markstur | bswartz, It _looked_ like Goutham changed that one -- but it turned out to be a gerrit UI trick for me | 21:18 |
markstur | So it seems that it usually does the right thing | 21:18 |
bswartz | I'm losing faith in the new gerrit and starting to miss the old gerrit | 21:19 |
markstur | bswartz, "starting to"??? You're late to that party. | 21:19 |
bswartz | well I've tried hard to be open minded about it | 21:20 |
bswartz | the new gerrit definitely has some upsides | 21:20 |
bswartz | but the downsides are accumulating | 21:21 |
gouthamr_ | i miss being able to use ctrl/cmd+F | 21:23 |
gouthamr_ | or seeing dependencies in an intuitive way | 21:25 |
*** eharney has quit IRC | 21:33 | |
*** eharney has joined #openstack-manila | 21:35 | |
*** eharney_ has joined #openstack-manila | 21:38 | |
*** eharney has quit IRC | 21:42 | |
*** mtanino has joined #openstack-manila | 21:43 | |
*** pcaruana has joined #openstack-manila | 21:48 | |
*** haomaiwang has joined #openstack-manila | 21:53 | |
*** haomaiwang has quit IRC | 21:57 | |
*** eharney_ has quit IRC | 22:05 | |
*** erlon has quit IRC | 22:06 | |
*** dustins has quit IRC | 22:14 | |
*** tpsilva has quit IRC | 22:16 | |
*** xyang_ has quit IRC | 22:17 | |
*** eharney_ has joined #openstack-manila | 22:18 | |
*** JoseMello has quit IRC | 22:18 | |
*** xyang1 has quit IRC | 22:21 | |
*** haomaiwang has joined #openstack-manila | 22:33 | |
*** haomaiwang has quit IRC | 22:38 | |
*** dcain1 has quit IRC | 22:49 | |
*** akerr_ is now known as akerr_away | 23:04 | |
*** akerr_away is now known as akerr_ | 23:06 | |
*** akerr_ is now known as akerr_away | 23:06 | |
*** akerr_away is now known as akerr_ | 23:06 | |
*** akerr_ is now known as akerr_away | 23:06 | |
*** akerr_away is now known as akerr_ | 23:07 | |
*** akerr_ is now known as akerr_away | 23:08 | |
*** akerr_away is now known as akerr_ | 23:08 | |
*** akerr_ is now known as akerr_away | 23:08 | |
*** akerr_away is now known as akerr_ | 23:08 | |
*** akerr_ is now known as akerr_away | 23:09 | |
*** eharney_ is now known as eharney | 23:15 | |
*** akerr_away is now known as akerr_ | 23:18 | |
*** lgreg has joined #openstack-manila | 23:22 | |
*** lgreg has left #openstack-manila | 23:22 | |
*** xyang_ has joined #openstack-manila | 23:24 | |
*** bill_az has quit IRC | 23:34 | |
*** a_ta has quit IRC | 23:36 | |
*** akerr_ has quit IRC | 23:44 | |
*** akerr has joined #openstack-manila | 23:50 | |
*** martyturner has joined #openstack-manila | 23:58 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!