*** alyson_ has joined #openstack-manila | 00:03 | |
*** Suyi_ has quit IRC | 00:05 | |
*** yangyapeng has quit IRC | 00:05 | |
openstackgerrit | Alyson proposed openstack/manila: Fix access rules for managed shares in HSP driver https://review.openstack.org/366310 | 00:06 |
---|---|---|
*** cknight has joined #openstack-manila | 00:09 | |
openstackgerrit | Alyson proposed openstack/manila: Fix access rules for managed shares in HSP driver https://review.openstack.org/366310 | 00:09 |
*** ociuhandu has quit IRC | 00:15 | |
*** MikeG451 has quit IRC | 00:17 | |
*** cknight has quit IRC | 00:18 | |
*** MikeG451 has joined #openstack-manila | 00:32 | |
*** kaisers1 has joined #openstack-manila | 00:32 | |
*** kaisers2 has joined #openstack-manila | 00:33 | |
*** kaisers has quit IRC | 00:34 | |
openstackgerrit | Alyson proposed openstack/manila: Add cleanup to create from snap in Manila HNAS driver https://review.openstack.org/356087 | 00:34 |
*** 32NABYQA1 has quit IRC | 00:35 | |
openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Improve Share Migration tempest tests https://review.openstack.org/366324 | 00:51 |
openstackgerrit | Jay Mehta proposed openstack/manila: 3PAR driver fails to validate conf share server IPs https://review.openstack.org/369763 | 00:53 |
*** gouthamr has quit IRC | 01:17 | |
openstackgerrit | Jay Mehta proposed openstack/manila: Correctly calling setup_server() calls in unit-tests https://review.openstack.org/360088 | 01:29 |
*** yangyapeng has joined #openstack-manila | 01:30 | |
openstackgerrit | Ha Van Tu proposed openstack/manila: [api-ref] Refactor consistency group API https://review.openstack.org/367711 | 01:40 |
*** zengyingzhe has quit IRC | 01:42 | |
openstackgerrit | Ha Van Tu proposed openstack/manila: [api-ref] Refactor consistency group API https://review.openstack.org/367711 | 01:50 |
*** esker has joined #openstack-manila | 01:50 | |
*** yangyapeng has quit IRC | 02:06 | |
*** yangyapeng has joined #openstack-manila | 02:07 | |
*** yangyape_ has joined #openstack-manila | 02:16 | |
*** yangyapeng has quit IRC | 02:18 | |
*** yangyapeng has joined #openstack-manila | 02:24 | |
*** yangyape_ has quit IRC | 02:27 | |
*** senk has quit IRC | 02:38 | |
*** senk has joined #openstack-manila | 02:38 | |
*** alyson_ has quit IRC | 02:39 | |
*** ociuhandu has joined #openstack-manila | 02:42 | |
*** akshai has joined #openstack-manila | 02:48 | |
*** akshai_ has joined #openstack-manila | 02:49 | |
*** akshai has quit IRC | 02:53 | |
*** senk has quit IRC | 03:09 | |
*** esker has quit IRC | 03:14 | |
*** esker has joined #openstack-manila | 03:14 | |
*** jcsp has quit IRC | 03:14 | |
*** akshai_ has quit IRC | 03:19 | |
*** esker is now known as esker[away] | 03:20 | |
*** yangyapeng has quit IRC | 03:31 | |
*** ganso has quit IRC | 03:55 | |
*** yangyapeng has joined #openstack-manila | 04:41 | |
*** kaisers has joined #openstack-manila | 04:43 | |
*** yangyapeng has quit IRC | 05:04 | |
*** yangyapeng has joined #openstack-manila | 05:06 | |
*** lpetrut has joined #openstack-manila | 05:18 | |
*** senk has joined #openstack-manila | 05:23 | |
*** nkrinner_afk is now known as nkrinner | 05:52 | |
*** Rajesh has joined #openstack-manila | 05:52 | |
Rajesh | hi | 05:52 |
*** Rajesh is now known as Guest69504 | 05:53 | |
*** JoseMello has joined #openstack-manila | 05:55 | |
Guest69504 | hey | 06:01 |
Guest69504 | I have a question .. | 06:01 |
openstackgerrit | zzxwill proposed openstack/manila: Put all imports from manila.i18n in one line https://review.openstack.org/361099 | 06:08 |
Guest69504 | I have a custom driver implemented for MapR Filesystem .. I want to push to openstack manila .. | 06:12 |
Guest69504 | can you let us know the procedure .. | 06:12 |
*** senk has quit IRC | 06:19 | |
*** nherciu has joined #openstack-manila | 06:22 | |
*** lpetrut has quit IRC | 06:28 | |
*** pcaruana has joined #openstack-manila | 06:32 | |
*** senk has joined #openstack-manila | 06:52 | |
*** senk has quit IRC | 07:01 | |
*** FL1SK has quit IRC | 07:02 | |
*** senk has joined #openstack-manila | 07:05 | |
zhongjun_ | Guest69504: Please see this link: https://github.com/openstack/manila/blob/master/doc/source/devref/driver_requirements.rst | 07:06 |
*** zengyingzhe has joined #openstack-manila | 07:09 | |
zhongjun_ | Guest69504: After you have finished all of the minimum requirements, and your patch have two or more +2, your driver code can be merged in openstack manila. | 07:12 |
zhongjun_ | Guest69504: In your MapR driver code patch: https://review.openstack.org/#/c/350035/, As Clinton Knight said, The MapR-FS driver met the deadline(link: https://openstack.nimeyo.com/93839/openstack-dev-manila-feature-proposal-freeze-is-here), but the associated CI did not, so this cannot merge in Newton. | 07:23 |
*** Guest69504 has quit IRC | 07:32 | |
*** aovchinnikov has joined #openstack-manila | 07:34 | |
*** dsariel has joined #openstack-manila | 07:55 | |
*** permalac has joined #openstack-manila | 08:03 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/manila-ui: Imported Translations from Zanata https://review.openstack.org/369890 | 08:17 |
*** akapil has joined #openstack-manila | 08:38 | |
*** jcsp has joined #openstack-manila | 08:44 | |
*** daidv has joined #openstack-manila | 08:47 | |
*** daidv has quit IRC | 08:51 | |
*** FL1SK has joined #openstack-manila | 09:02 | |
*** daidv has joined #openstack-manila | 09:15 | |
*** akapil has quit IRC | 09:19 | |
*** akapil has joined #openstack-manila | 09:21 | |
*** akapil has quit IRC | 09:25 | |
*** akapil has joined #openstack-manila | 09:29 | |
*** permalac has quit IRC | 09:38 | |
*** JoseMello has quit IRC | 09:51 | |
*** akapil has quit IRC | 09:52 | |
*** yangyapeng has quit IRC | 09:53 | |
*** lpetrut has joined #openstack-manila | 09:57 | |
*** cknight has joined #openstack-manila | 10:00 | |
*** jcsp has quit IRC | 10:06 | |
*** cknight has quit IRC | 10:10 | |
*** cknight has joined #openstack-manila | 10:20 | |
*** rraja has joined #openstack-manila | 10:21 | |
*** cknight has quit IRC | 10:35 | |
*** senk has quit IRC | 10:36 | |
*** alyson_ has joined #openstack-manila | 10:55 | |
*** senk has joined #openstack-manila | 11:00 | |
*** ganso has joined #openstack-manila | 11:10 | |
*** daidv has quit IRC | 11:22 | |
*** openstackgerrit has quit IRC | 11:34 | |
*** openstackgerrit has joined #openstack-manila | 11:35 | |
*** daidv has joined #openstack-manila | 11:56 | |
tbarron | zhongjun_: did the translations get corrected? | 12:09 |
*** gouthamr has joined #openstack-manila | 12:10 | |
tbarron | zhongjun_: in manila-ui, and thank you for reviewing! | 12:10 |
*** esker[away] is now known as esker | 12:13 | |
*** esker has quit IRC | 12:29 | |
*** esker has joined #openstack-manila | 12:30 | |
openstackgerrit | Valeriy Ponomaryov proposed openstack/manila: [ZFSonLinux] Fix share migration using remote host https://review.openstack.org/370095 | 12:33 |
*** esker has quit IRC | 12:34 | |
ganso | vponomaryov: ping | 12:36 |
vponomaryov | ganso: pong | 12:36 |
ganso | vponomaryov: Hi Valeriy | 12:36 |
vponomaryov | ganso: Hello Rodrigo ) | 12:37 |
ganso | vponomaryov: it seems the problem was not firewall related, in https://review.openstack.org/#/c/365774 | 12:37 |
ganso | vponomaryov: have you been able to achieve host-assisted migration in your lab? | 12:37 |
vponomaryov | ganso: i didn't try | 12:37 |
*** xyang1 has joined #openstack-manila | 12:37 | |
vponomaryov | ganso: only driver-assisted | 12:37 |
ganso | vponomaryov: since you have it set up, would you kindly try the host-assisted approach to see what error pops up? | 12:39 |
vponomaryov | ganso: you so sure it will fail? ) | 12:40 |
ganso | vponomaryov: I am not, but if it does work, then I am sure it does not need a bugfix | 12:40 |
ganso | vponomaryov: just CI configuration | 12:40 |
*** ociuhandu has quit IRC | 12:42 | |
*** lpetrut has quit IRC | 12:43 | |
*** lpetrut has joined #openstack-manila | 12:43 | |
*** dsariel has quit IRC | 12:49 | |
vponomaryov | ganso: I get another error: http://paste.ofcode.org/GKVEXkdKFpMYySdtBRXEge | 12:50 |
*** porrua has joined #openstack-manila | 12:54 | |
ganso | vponomaryov: humm | 13:00 |
ganso | vponomaryov: says timeout, permission denied | 13:00 |
ganso | vponomaryov: isn't it running exactly the same command you run on your host to mount the share on it? | 13:00 |
vponomaryov | ganso: yes | 13:03 |
vponomaryov | ganso: except dst dir, that I create manually | 13:03 |
ganso | vponomaryov: do you use 'manila access-allow share_x ip 127.0.0.1' as well or another ip ? | 13:03 |
vponomaryov | yes | 13:03 |
ganso | vponomaryov: yea dest dir is /tmp/share_x in the data service, but that should not prevent it from working | 13:04 |
ganso | vponomaryov: this is very weird | 13:05 |
*** cknight has joined #openstack-manila | 13:07 | |
*** StraubTW has joined #openstack-manila | 13:09 | |
*** daidv has quit IRC | 13:17 | |
*** cknight1 has joined #openstack-manila | 13:19 | |
openstackgerrit | Rodrigo Barbieri proposed openstack/manila: Add cleanup to create from snap in Manila HNAS driver https://review.openstack.org/356087 | 13:20 |
*** cknight has quit IRC | 13:21 | |
*** dsariel has joined #openstack-manila | 13:26 | |
*** dsariel has quit IRC | 13:34 | |
*** zhugaoxiao has quit IRC | 13:35 | |
*** dsariel has joined #openstack-manila | 13:40 | |
*** akerr has joined #openstack-manila | 13:43 | |
openstackgerrit | Merged openstack/manila: huawei driver default create thin type share https://review.openstack.org/351475 | 13:46 |
*** akerr has quit IRC | 13:48 | |
*** mtanino has joined #openstack-manila | 13:49 | |
*** akerr has joined #openstack-manila | 13:49 | |
*** rooneym has joined #openstack-manila | 13:59 | |
*** akshai has joined #openstack-manila | 14:01 | |
*** eharney has joined #openstack-manila | 14:07 | |
*** dustins has joined #openstack-manila | 14:07 | |
*** dsariel has quit IRC | 14:09 | |
*** cknight1 has quit IRC | 14:14 | |
*** vbellur has quit IRC | 14:26 | |
*** daidv has joined #openstack-manila | 14:33 | |
*** daidv has quit IRC | 14:33 | |
*** nherciu has quit IRC | 14:36 | |
*** nkrinner is now known as nkrinner_afk | 14:37 | |
*** Yogi1 has joined #openstack-manila | 14:41 | |
*** vbellur has joined #openstack-manila | 14:50 | |
*** erlon has joined #openstack-manila | 14:50 | |
*** daidv has joined #openstack-manila | 14:50 | |
*** daidv has quit IRC | 14:50 | |
*** dsariel has joined #openstack-manila | 14:53 | |
*** senk has quit IRC | 14:56 | |
*** timcl has quit IRC | 14:58 | |
*** timcl has joined #openstack-manila | 14:59 | |
*** akshai has quit IRC | 14:59 | |
*** akshai has joined #openstack-manila | 15:01 | |
openstackgerrit | Merged openstack/manila: Put all imports from manila.i18n in one line https://review.openstack.org/361099 | 15:04 |
*** dsariel has quit IRC | 15:17 | |
*** aovchinnikov has quit IRC | 15:20 | |
openstackgerrit | Doug Hellmann proposed openstack/manila-ui: Update reno for stable/newton https://review.openstack.org/370207 | 15:21 |
*** senk has joined #openstack-manila | 15:27 | |
*** senk has quit IRC | 15:28 | |
*** pcaruana has quit IRC | 15:40 | |
*** faiz89 has joined #openstack-manila | 15:42 | |
*** absubram has joined #openstack-manila | 15:51 | |
*** faiz89 is now known as faiz89_away | 15:51 | |
*** Yogi1 has quit IRC | 15:51 | |
*** gouthamr-afk has joined #openstack-manila | 16:01 | |
*** gouthamr-afk is now known as gouthamr_ | 16:02 | |
*** gouthamr-afk has joined #openstack-manila | 16:02 | |
*** adrianofr_ has joined #openstack-manila | 16:04 | |
*** gouthamr_ has quit IRC | 16:07 | |
*** dsariel has joined #openstack-manila | 16:17 | |
*** vbellur has quit IRC | 16:20 | |
*** porrua has quit IRC | 16:29 | |
ganso | xyang1, gouthamr, bswartz, markstur, tbarron: Hello, could you please review https://review.openstack.org/#/c/365136/ when you have some time? Thanks in advance | 16:31 |
*** cknight has joined #openstack-manila | 16:31 | |
*** vbellur has joined #openstack-manila | 16:35 | |
*** gouthamr_afk has joined #openstack-manila | 16:37 | |
*** gouthamr-afk has quit IRC | 16:40 | |
*** cknight has quit IRC | 16:41 | |
*** cknight1 has joined #openstack-manila | 16:42 | |
openstackgerrit | Alyson proposed openstack/manila: Add cleanup to create from snap in Manila HNAS driver https://review.openstack.org/356087 | 16:46 |
*** lpetrut has quit IRC | 16:47 | |
*** dsariel has quit IRC | 16:47 | |
*** porrua has joined #openstack-manila | 16:58 | |
*** faiz89_away is now known as faiz89 | 17:12 | |
*** Suyi_ has joined #openstack-manila | 17:20 | |
*** dsariel has joined #openstack-manila | 17:22 | |
gouthamr | ganso: ping | 17:32 |
ganso | gouthamr: pong | 17:32 |
gouthamr | ganso: Hey! so, i was thinking about the sanest way to avoid those RPC requests from becoming driver calls.. i think we need a "deleting" state as well to recurse upon, so as to safely remove the lock around the gigantic critical section and place the lock only around DB reads and writes. | 17:33 |
ganso | gouthamr: did you see my comment? one of my suggestions is that "deleting" status | 17:34 |
gouthamr | ganso: yes :) i'm not comfortable making that change for RC1 however, i discussed with bswartz and he thinks adding this DB migration isn't the safest thing for RC1 too.do you see any compelling reasons we can't keep existing behavior for Newton.. | 17:35 |
gouthamr | looks like the bug was also re-targeted. | 17:36 |
ganso | gouthamr: I agree with bswartz, I was very worried before because I thought that when the API got stuck there was no workaround, but I found that if you delete a rule it fixes it. So I think we are safe and we should not include this drastic change | 17:37 |
gouthamr | ganso: yes.. there's a way to "recover" from lost access rule updates | 17:37 |
gouthamr | ganso: but this refactor makes sense for Ocata and beyond. | 17:38 |
ganso | gouthamr: yes, makes a lot of sense, I like the direction it is going | 17:38 |
gouthamr | ganso: thanks! will update the patch with some more changes.. | 17:39 |
*** rhagarty has joined #openstack-manila | 17:39 | |
*** dsariel has quit IRC | 17:42 | |
tbarron | oslo has flavio percocco, julien danjou, victor stinner, ben nemec | 17:58 |
tbarron | sorry, wrong channel :) | 18:07 |
gouthamr | tbarron: :P this is the only channel | 18:07 |
tbarron | gouthamr: my heart is here, but my eyes roam | 18:08 |
gouthamr | tbarron: i hear you.. :D | 18:08 |
* bswartz lurks in 25 freenode channels | 18:08 | |
openstackgerrit | Jay Mehta proposed openstack/manila: 3PAR driver fails to validate conf share server IPs https://review.openstack.org/369763 | 18:10 |
*** faiz89 is now known as faiz89_away | 18:31 | |
*** rraja has quit IRC | 18:33 | |
*** faiz89_away is now known as faiz89 | 18:37 | |
*** lpetrut has joined #openstack-manila | 18:40 | |
*** akerr has quit IRC | 18:43 | |
*** akerr has joined #openstack-manila | 18:46 | |
*** lpetrut has quit IRC | 18:47 | |
*** lpetrut has joined #openstack-manila | 18:49 | |
*** assassin has quit IRC | 18:51 | |
*** absubram has quit IRC | 19:03 | |
*** pcaruana has joined #openstack-manila | 19:07 | |
ganso | xyang1, bswartz, cknight1, markstur, tbarron: CIs have run on https://review.openstack.org/#/c/356087/ and https://review.openstack.org/#/c/366310 | 19:10 |
*** faiz89 is now known as faiz89_away | 19:16 | |
*** faiz89_away is now known as faiz89 | 19:16 | |
*** faiz89 is now known as faiz89_away | 19:16 | |
*** faiz89_away has quit IRC | 19:17 | |
*** kaisers has quit IRC | 19:22 | |
ganso | cknight1, tbarron, vponomaryov: thanks! | 19:23 |
alyson_ | ^ thanks :) | 19:25 |
*** akerr has quit IRC | 19:25 | |
*** akshai has quit IRC | 19:28 | |
*** lpetrut has quit IRC | 19:32 | |
*** akerr has joined #openstack-manila | 19:34 | |
*** gouthamr_afk has quit IRC | 19:35 | |
*** lpetrut has joined #openstack-manila | 19:36 | |
StraubTW | Does anyone have an idea of which parameter I am missing if I get an AmbiguousEndpoints error when I try to create a share? | 19:39 |
StraubTW | Exception during message handling: AmbiguousEndpoints: [{u'adminURL': u'http://dev01-admin.os.cloud.twc.net:8774/v2/5fd36947639047259043575aeeda84a9', u'region': u'dev01', u'internalURL': u'http://dev01-internal.os.cloud.twc.net:8774/v2/5fd36947639047259043575aeeda84a9', 'serviceName': u'nova', u'id': u'8e681e46ed35445082053a81ef555eb2', u'publicURL': u'http://dev01.os.cloud.twc.net:8774/v2/5fd3694763904725 | 19:39 |
StraubTW | 9043575aeeda84a9'}, {u'adminURL': u'http://dev02-admin.os.cloud.twc.net:8774/v2/5fd36947639047259043575aeeda84a9', u'region': u'dev02', u'internalURL': u'http://dev02-internal.os.cloud.twc.net:8774/v2/5fd36947639047259043575aeeda84a9', 'serviceName': u'nova', u'id': u'92af57b54250435a9047092a7f31992d', u'publicURL': u'http://dev02.os.cloud.twc.net:8774/v2/5fd36947639047259043575aeeda84a9'}] | 19:39 |
StraubTW | My regions are dev02 and dev01. I have manila running on dev02. | 19:39 |
*** lpetrut has quit IRC | 19:40 | |
StraubTW | os_region_name = dev02 and region_name = dev02 in manila.conf | 19:40 |
StraubTW | Keystone is returning two endpoints from the manila/compute/nova.py call | 19:42 |
StraubTW | that goes to novaclient service_catalog.py | 19:43 |
*** akshai has joined #openstack-manila | 19:45 | |
*** akshai_ has joined #openstack-manila | 19:46 | |
*** porrua has quit IRC | 19:48 | |
*** akshai has quit IRC | 19:49 | |
*** hoonetorg has quit IRC | 19:53 | |
openstackgerrit | Merged openstack/manila: Fix access rules for managed shares in HSP driver https://review.openstack.org/366310 | 19:56 |
*** akerr has quit IRC | 19:57 | |
ganso | vponomaryov: ping | 19:58 |
*** pcaruana has quit IRC | 20:00 | |
*** faiz89 has joined #openstack-manila | 20:00 | |
*** vbellur has quit IRC | 20:01 | |
bswartz | ganso: how come you keep targeting more bugs to rc1? | 20:01 |
bswartz | we need less bugs not more | 20:01 |
*** rhagarty has quit IRC | 20:02 | |
openstackgerrit | Merged openstack/manila: Add cleanup to create from snap in Manila HNAS driver https://review.openstack.org/356087 | 20:02 |
ganso | bswartz: it is the one we discussed yesterday, you asked me if we really need it | 20:02 |
ganso | bswartz: also, it is the almost the same as the one triaged, but with a proposed solution for now | 20:02 |
bswartz | oh | 20:02 |
ganso | bswartz: the triaged should be removed | 20:02 |
bswartz | I see | 20:03 |
ganso | bswartz: from the RC1 target | 20:03 |
bswartz | yeah I thought that issue has been targeted for a while and I didn't realize this bug was related | 20:03 |
bswartz | did we really need a second bug? | 20:03 |
*** rhagarty has joined #openstack-manila | 20:04 | |
ganso | bswartz: yes, because the previous bug was expecting a different behavior to be correct | 20:04 |
ganso | bswartz: previous bug *entry | 20:05 |
bswartz | okay | 20:05 |
bswartz | so I can untarget https://bugs.launchpad.net/manila/+bug/1619788 then?" | 20:05 |
openstack | Launchpad bug 1619788 in Manila "Share migration does not allow_access before migration_complete" [High,Triaged] - Assigned to Rodrigo Barbieri (rodrigo-barbieri2010) | 20:05 |
ganso | bswartz: yes | 20:05 |
ganso | bswartz: the possible fix for that one we decided at monday's discussion to postpone to Ocata | 20:05 |
*** gouthamr has quit IRC | 20:07 | |
*** hoonetorg has joined #openstack-manila | 20:08 | |
bswartz | ganso: what is MANILA_DATA_COPY_CHECK_HASH? | 20:08 |
bswartz | oh it's a m-dat option | 20:09 |
ganso | bswartz: it was a config that enables the use of hash validation for all data copies in Data Service. We did not have an option for it before, it was disabled, I decided to add an option for it so it could be tested in gate | 20:09 |
bswartz | yeah it makes sense | 20:09 |
ganso | bswartz: having a code that is not tested in gate was not ideal | 20:09 |
bswartz | I'm good with this last bugfix then | 20:09 |
bswartz | 3 remaining bugs | 20:10 |
bswartz | 2 have workflow | 20:10 |
ganso | bswartz: all we need now is another +2 for https://review.openstack.org/#/c/365136 | 20:10 |
bswartz | yes | 20:10 |
ganso | bswartz: also, we should merge the spec https://review.openstack.org/#/c/315707 | 20:11 |
bswartz | markstur xyang1 vponomaryov tbarron toabctl: last bug for RC1 https://review.openstack.org/#/c/365136 | 20:11 |
ganso | bswartz: it is up-to-date | 20:11 |
bswartz | ninja'd | 20:12 |
ganso | bswartz: alright | 20:13 |
openstackgerrit | Merged openstack/manila-specs: Add spec for Share Migration Newton improvements https://review.openstack.org/315707 | 20:18 |
*** vbellur has joined #openstack-manila | 20:22 | |
*** vbellur has quit IRC | 20:28 | |
*** vbellur has joined #openstack-manila | 20:28 | |
*** vbellur has quit IRC | 20:38 | |
*** gouthamr has joined #openstack-manila | 20:44 | |
*** absubram has joined #openstack-manila | 20:49 | |
*** akshai_ has quit IRC | 20:51 | |
*** cknight1 has quit IRC | 20:51 | |
*** faiz89 has quit IRC | 20:54 | |
openstackgerrit | Merged openstack/manila: [ZFSonLinux] Fix share migration using remote host https://review.openstack.org/370095 | 20:55 |
*** ubijtsa has joined #openstack-manila | 21:07 | |
*** ubijtsa is now known as Guest25001 | 21:08 | |
openstackgerrit | Merged openstack/manila: Improve Share Migration tempest tests https://review.openstack.org/366324 | 21:08 |
openstackgerrit | Merged openstack/manila: Fix allow/deny error message and race in migration https://review.openstack.org/365136 | 21:13 |
*** vbellur has joined #openstack-manila | 21:17 | |
*** daidv has joined #openstack-manila | 21:18 | |
*** daidv has quit IRC | 21:18 | |
*** faiz89 has joined #openstack-manila | 21:31 | |
*** rhagarty has quit IRC | 21:39 | |
*** vbellur has quit IRC | 21:54 | |
*** vbellur has joined #openstack-manila | 21:54 | |
*** lpetrut has joined #openstack-manila | 22:06 | |
*** alyson_ has quit IRC | 22:09 | |
*** StraubTW has quit IRC | 22:09 | |
*** eharney has quit IRC | 22:11 | |
gouthamr | ganso: here? | 22:15 |
*** faiz89 has quit IRC | 22:16 | |
ganso | gouthamr: yea | 22:16 |
gouthamr | ganso: hey Rodrigo.. i've a question if you have a minute | 22:17 |
ganso | gouthamr: sure | 22:17 |
gouthamr | ganso: with update_access being an interface to add/remove rules.. with the per-rule-per-share-instance status back, i see no reason to have two separate RPC methods; i.e, allow_access and deny_access | 22:18 |
ganso | gouthamr: in share/manager, right? | 22:18 |
gouthamr | ganso: if we set the rule statuses to 'new' or 'deleting'; we can just invoke update_access to do the right thing | 22:18 |
*** rooneym has quit IRC | 22:19 | |
gouthamr | ganso: yes.. not talking about the driver methods, just the share/rpcapi | 22:19 |
ganso | gouthamr: yes :) | 22:19 |
*** lpetrut has quit IRC | 22:19 | |
ganso | gouthamr: we will still need share/api allow and deny | 22:19 |
ganso | gouthamr: but both of them invoke update_access just as a refresh | 22:19 |
gouthamr | ganso: yes.. | 22:19 |
ganso | gouthamr: or, if we go the periodic task approach, no need for RPCs | 22:20 |
gouthamr | ganso: i thought about that.. but maybe you don't want the access check to run periodically; especially if drivers are doing a lot of processing | 22:20 |
ganso | gouthamr: yes, we would also need to be extra careful with statuses | 22:21 |
gouthamr | ganso: like with everything else :P | 22:22 |
gouthamr | ganso: i realized i need to switch all "applying" to "new" on driver restarts. | 22:23 |
ganso | gouthamr: for 2 reasons: 1) any method anywhere may trigger update_access through the periodic task just by adding/removing rules, changing statuses. 2) it may trigger update_access in the middle of a operation that is changing the status, like last steps of migration | 22:23 |
*** mtanino has quit IRC | 22:23 | |
ganso | gouthamr: good catch! | 22:23 |
ganso | gouthamr: we need to remember driver restarts | 22:23 |
ganso | gouthamr: so, with the periodic task it may go out of control | 22:24 |
gouthamr | ganso: yes.. | 22:24 |
ganso | gouthamr: plus the added latency | 22:24 |
ganso | gouthamr: also, are you still wanting to remove the lock? | 22:24 |
gouthamr | ganso: true.. all periodic tasks in the manager should ideally be configurable; so i doubt we need one more configuration headache :P | 22:25 |
gouthamr | ganso: yeah, going to think about that... we need a read/write lock on the state change instead of locking the whole driver invocation.. | 22:27 |
ganso | gouthamr: it is not the state change that require lock anymore | 22:28 |
gouthamr | ganso: at this point though, i think we can safely consume the RPC call and not call the access helper if it's already updating | 22:28 |
ganso | gouthamr: it is the resource and serialization | 22:28 |
ganso | gouthamr: how can you determine it is already updating? checking the state will introduce a race | 22:29 |
gouthamr | ganso: access helper can keep track of what share instances are currently being updated through it | 22:29 |
gouthamr | ganso: so, a helper method to push/pop into that list of share instances can be protected by a lock. | 22:30 |
gouthamr | ganso: to avoid any races in that tiny gap | 22:30 |
ganso | gouthamr: yes, a locked object | 22:30 |
ganso | gouthamr: nice | 22:31 |
gouthamr | ganso: good stuff.. thanks.. | 22:31 |
*** adrianofr_ has quit IRC | 22:32 | |
*** ociuhandu has joined #openstack-manila | 22:33 | |
ganso | gouthamr: you're welcome! ping me when you have an update, I'll take a look and help test :) | 22:33 |
gouthamr | ganso: sure thing | 22:33 |
*** absubram has quit IRC | 22:57 | |
*** ChanServ changes topic to "OpenStack Shared File Systems | Manila | Newton RC1 complete! Keep testing and looking for critical bugs." | 22:59 | |
*** dustins has quit IRC | 23:00 | |
*** absubram has joined #openstack-manila | 23:01 | |
*** xyang1 has quit IRC | 23:10 | |
*** Suyi_ has quit IRC | 23:25 | |
*** ociuhandu has quit IRC | 23:49 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!