| *** EmilDi has quit IRC | 00:33 | |
| *** dschroeder has quit IRC | 00:39 | |
| *** yangyapeng has joined #openstack-freezer | 01:01 | |
| *** yangyapeng has quit IRC | 01:02 | |
| *** EinstCrazy has joined #openstack-freezer | 01:06 | |
| *** yangyapeng has joined #openstack-freezer | 01:10 | |
| *** jith_ has quit IRC | 02:32 | |
| *** jith_ has joined #openstack-freezer | 04:47 | |
| jith_ | Hi all, is freezer-api required in between backingup in ssh location and freezer-scheduler... | 04:51 |
|---|---|---|
| jith_ | means freezer scheduler will work only with freezer-api?? because scheduler uploads job to api... | 04:53 |
| *** dmellado has quit IRC | 06:34 | |
| *** dmellado has joined #openstack-freezer | 06:39 | |
| *** jith_ has quit IRC | 07:40 | |
| *** daemontool__ has joined #openstack-freezer | 08:38 | |
| *** daemontool_ has quit IRC | 08:41 | |
| *** openstackgerrit has quit IRC | 08:48 | |
| *** openstackgerrit has joined #openstack-freezer | 08:48 | |
| *** jith_ has joined #openstack-freezer | 09:02 | |
| openstackgerrit | Fabrizio Vanni proposed openstack/freezer: enable output of metadata to a file https://review.openstack.org/284629 | 09:25 |
| *** yangyapeng has quit IRC | 09:40 | |
| *** yangyapeng has joined #openstack-freezer | 09:40 | |
| *** EinstCrazy has quit IRC | 09:54 | |
| daemontool__ | jith_, hi | 09:54 |
| daemontool__ | regarding your questions | 09:54 |
| *** EinstCrazy has joined #openstack-freezer | 09:54 | |
| daemontool__ | the freezer-api are not required to take backup and store them to an ssh node | 09:55 |
| daemontool__ | the scheduler works also without api | 09:55 |
| daemontool__ | there's an option if I remembe rwell called --no-api | 09:55 |
| jith_ | daemontool_: thanks | 09:56 |
| daemontool__ | jith_, to take backups and store them to an ssh node | 09:59 |
| daemontool__ | you do not even need the scheduler if you do not want it for some reason | 09:59 |
| *** EinstCrazy has quit IRC | 10:15 | |
| *** EinstCrazy has joined #openstack-freezer | 10:16 | |
| *** EinstCrazy has quit IRC | 10:21 | |
| *** yangyapeng has quit IRC | 10:23 | |
| *** reldan has joined #openstack-freezer | 10:32 | |
| *** reldan has left #openstack-freezer | 10:36 | |
| *** daemontool has joined #openstack-freezer | 10:36 | |
| jith_ | daemontool_ :thanks.. sorry i was away | 10:39 |
| *** daemontool__ has quit IRC | 10:40 | |
| jith_ | ok but i want to take daily backups.. so scheduler is needed right? | 10:40 |
| *** reldan has joined #openstack-freezer | 10:42 | |
| vannif_ | the scheduler-api infrastructure provide some features to the backup service: a single endpoint to interact with to distribute and monitor the execution of the backups on different clients or nodes. | 10:43 |
| vannif_ | if the problem is really small (single node with recurring backups) you could use a simple cron job to execute the freezer-agent. | 10:45 |
| jith_ | so if i am enabling freezer in openstack, freezer-api and freezer should be in controller as well as compute nodes??? | 10:45 |
| jith_ | vannif: thanks i got | 10:45 |
| vannif_ | you're welcome. what is your objective ? single node ? multiple VM ? backup of OpenStack control plane ? | 10:46 |
| jith_ | backup of openstack control plane | 10:47 |
| jith_ | as of now.. i am trying to execute in single node | 10:50 |
| jith_ | scheduler should be in all nodes or only controller node? | 10:51 |
| reldan | Please review https://review.openstack.org/#/c/284637/ | 10:55 |
| vannif_ | https://review.openstack.org/#/c/284629/ | 11:02 |
| m3m0 | vannif_ done | 11:08 |
| m3m0 | reldan just a quick question in the patch | 11:08 |
| reldan | yes | 11:08 |
| m3m0 | i wrote the question in the patch | 11:09 |
| reldan | m3m0: nope currently we have 2 places with defaults | 11:10 |
| reldan | 1) cli 2) dictionary called DEFAULT_PARAMS | 11:10 |
| reldan | and it just has no sense | 11:10 |
| reldan | so if you don’t provide mode, mode should be taken from dict | 11:11 |
| reldan | m3m0: 'mode': 'fs' | 11:11 |
| m3m0 | ok ok make sense | 11:14 |
| reldan | m3m0: Thank you! | 11:15 |
| vannif_ | jith_: the scheduler is the process that contacts the api, fetches the jobs and passes the instructions to the freezer-agent to execute the backups (or whatever action is written in the job document). You need to run it on any node that needs to execute the freezer-agent | 11:16 |
| openstackgerrit | Memo Garcia proposed openstack/freezer-web-ui: Improve success messages when it takes time to display the action https://review.openstack.org/286507 | 11:16 |
| m3m0 | slashme, did we provide the test environment to the QA guys? | 11:17 |
| jith_ | vannif_ thanks, i got that.. so scheduler should run in all nodes?? | 11:18 |
| daemontool | all, FYI https://review.openstack.org/#/c/271072/ | 11:19 |
| daemontool | jith_, the scheduler run on any node where the data you want to backup is | 11:20 |
| daemontool | so yes, in theory on any node | 11:20 |
| jith_ | so if i want to take a backup of a nova instance, i should upload the scheduler job from the particular compute node? | 11:20 |
| daemontool | if you want to backup a nova instance, it is not required as currently we use the nova api | 11:21 |
| daemontool | so the scheduler can run on the instance itself | 11:21 |
| daemontool | the concept is> if you have to backup files on the file system then the scheduler needs to run on the node where the data is | 11:21 |
| daemontool | if you have to backup volumes and vms, then you can run the scheduler/agent where you want | 11:22 |
| jith_ | fine.. cant i take backup of instance from compute node?? | 11:22 |
| daemontool | yes | 11:22 |
| daemontool | ah ok | 11:22 |
| daemontool | do you mean | 11:22 |
| daemontool | directly from the file system like /var/lib/nova/instances? | 11:22 |
| daemontool | rather than using the nova api? something liek that? | 11:23 |
| jith_ | daemontool: yes yes | 11:23 |
| daemontool | jith_, good question | 11:23 |
| daemontool | we are working on that | 11:23 |
| daemontool | not currenlty | 11:23 |
| daemontool | I mean | 11:23 |
| daemontool | you can | 11:23 |
| daemontool | the problem is that if you backup instances from /var/lib/nova/instances | 11:23 |
| daemontool | and then restore | 11:23 |
| daemontool | the state of the vms can be inconsistent with the nova db in mysql | 11:24 |
| daemontool | :( | 11:24 |
| daemontool | so we are working on delivering a feature that will solve that | 11:24 |
| daemontool | reldan, is the man | 11:24 |
| daemontool | frescof, ^^ | 11:24 |
| vannif_ | regardin the creation of the the job: it doesn't need to be created on the same node. in fact the common usage is to have the scheduler running on the respective nodes, while the jobs are created, uploaded, managed from a managment point, for example a admin node or using the horizon dashboard which is under development as well | 11:25 |
| jith_ | daemontool: ok fine... also i saw a line in the freezer doc | 11:25 |
| jith_ | Execute a nova backup: | 11:25 |
| jith_ | $ freezerc --nova-inst-id 3ad7a62f-217a-48cd-a861-43ec0a04a78b | 11:25 |
| daemontool | jith_, so we need to provide a mechanism that backup the instance incrementally, save the related metadata | 11:25 |
| daemontool | and then restore it | 11:26 |
| daemontool | yes | 11:26 |
| vannif_ | yes. if you backup the ephemeral disk, it's not even granted that the VM can boot. you're backin up a possibly inconsistend filesystem | 11:26 |
| jith_ | fine.. then for cinder volume?? | 11:27 |
| jith_ | Execute a cinder backup: | 11:27 |
| jith_ | $ freezerc --cinder-vol-id 3ad7a62f-217a-48cd-a861-43ec0a04a78b | 11:27 |
| daemontool | jith_, that just wrap this nova api http://developer.openstack.org/api-ref-compute-v2.1.html#createBackup | 11:28 |
| daemontool | unfortunately it does not provide incrementals | 11:28 |
| daemontool | atm | 11:28 |
| jith_ | daemontool: oh fine.. thanks | 11:29 |
| jith_ | cinder volume backup is incremental?? | 11:29 |
| daemontool | jith_, yes | 11:30 |
| daemontool | but it does the same | 11:30 |
| daemontool | it uses the cinder incremntals api | 11:30 |
| daemontool | and that is from liberty | 11:30 |
| jith_ | backup service? | 11:30 |
| daemontool | currently the limitation is | 11:31 |
| jith_ | cinder backup service? | 11:31 |
| daemontool | yes | 11:31 |
| daemontool | the limitation is that you can't backup volumes on an openstack instance and restore them on an independent openstack isntance | 11:31 |
| daemontool | cause we need to backup also the metadata to do that | 11:31 |
| jith_ | ok thanks for responding to my questions patiently | 11:31 |
| daemontool | also currently the backups are not removed using cinder backups | 11:32 |
| jith_ | yes | 11:32 |
| daemontool | I mean removed automatically | 11:32 |
| daemontool | that should be also delivered in Mitaka | 11:32 |
| jith_ | ok i got | 11:32 |
| daemontool | reldan, any thoughts/says on this? | 11:32 |
| daemontool | :) | 11:32 |
| reldan | daemontool: I actually don’t know what to say. Currently nova/cinder backup in freezer - it is just a calls of api | 11:33 |
| reldan | daemontool: nothing actually more | 11:33 |
| daemontool | ok | 11:33 |
| daemontool | reldan, when you can, please have a conversation with frescof | 11:34 |
| daemontool | he found an easy way | 11:34 |
| daemontool | to execute incrementals backup | 11:34 |
| daemontool | for vms | 11:34 |
| daemontool | (it requires rsync tho) | 11:34 |
| daemontool | but until that conversation&bp happens | 11:34 |
| daemontool | we can probably keep working on improving the volumes backups? | 11:34 |
| daemontool | like adding metadata backup | 11:35 |
| daemontool | and managing backup removal/retention? | 11:35 |
| reldan | daemontool: Yes, I would like to have something like architecture. What we are going to have for cinder/nova etc. | 11:35 |
| reldan | daemontool: And what we are not going to support | 11:35 |
| reldan | daemontool: Like would we want to support aws in feature? | 11:36 |
| reldan | daemontool: Or only openstack nova/volume | 11:36 |
| reldan | daemontool: or vmware | 11:36 |
| daemontool | ok | 11:36 |
| reldan | daemontool: If we want to support different cloud providers - we should think about architecture. How to isolate layers | 11:37 |
| daemontool | reldan, I agree | 11:37 |
| daemontool | the thing is that we do not have any use case for that... | 11:37 |
| daemontool | it's a great idea | 11:38 |
| reldan | daemontool: if we have no plans to support other cloud providers - we can just write cinder specific code and it is all | 11:38 |
| daemontool | we can do it | 11:38 |
| daemontool | but no one asked for it | 11:38 |
| daemontool | we can think abou tit | 11:38 |
| daemontool | what we really need is to provide users to manage volumes and vm instances incremental backups | 11:39 |
| reldan | about incremental nova - it would be great to read somewhere about this. like a blueprint or just several sentences in etherpad should be enough | 11:39 |
| daemontool | with automatic deletion/retention | 11:39 |
| daemontool | frescof, is writing that | 11:39 |
| reldan | great! | 11:39 |
| daemontool | as soon as he finish to add the info to the bp | 11:39 |
| daemontool | he should have written the bp | 11:39 |
| daemontool | yesterday | 11:39 |
| daemontool | for dr | 11:39 |
| daemontool | once we provide that features... | 11:40 |
| daemontool | I think we can make better as much as we want | 11:40 |
| daemontool | but let's think abou tit | 11:40 |
| *** reldan has quit IRC | 11:45 | |
| jith_ | freezer-api is needed in all nodes??.. i saw in architecture, API runs in controller node... but how come scheduler know API's credential... the below command should run in controller ??? | 11:45 |
| jith_ | upload that job into the API: | 11:45 |
| jith_ | freezer-scheduler -c node12 job-create --file test_job.json | 11:45 |
| jith_ | -c --> will be client node where backup files reside | 11:46 |
| jith_ | i am sorry if i m confusing... | 11:46 |
| daemontool | jith_, that are the normal credentials you have with OS_* env variables | 11:47 |
| daemontool | once the freezer-api are registered to keystone | 11:47 |
| daemontool | the api endpoint will be available | 11:47 |
| daemontool | from the keystone catalog | 11:47 |
| daemontool | to the scheduler authenticate to keystone (using the OS_* env var) | 11:48 |
| daemontool | s/to/so/ | 11:48 |
| jith_ | oh yes i forgot that.. | 11:48 |
| daemontool | then retrieve the freezer-api endpoint | 11:48 |
| daemontool | ok | 11:48 |
| daemontool | :) | 11:48 |
| daemontool | lol | 11:48 |
| jith_ | client_id is?? its own hostname?? | 11:49 |
| daemontool | it should be tenant_id + hostname | 11:53 |
| daemontool | vannif_, does the client_id is generated with OS_PROJECT_ID if OS_TENANT_ID is not found? | 11:55 |
| daemontool | jith_, do you mind add all these information you asked to the openstack/freezer/README in the related sectino? | 11:56 |
| daemontool | :) | 11:56 |
| jith_ | daemontool: thanks: we should specify the parameter right while executing scheduler??? | 11:59 |
| jith_ | never.. :)..but how can i add?? | 12:01 |
| daemontool | jith_, clone the freezer repo | 12:01 |
| daemontool | edit the README file | 12:02 |
| daemontool | make the docs better by providing these information | 12:02 |
| daemontool | and send it for review | 12:02 |
| jith_ | sure.. | 12:02 |
| daemontool | do you know how to use git? | 12:02 |
| jith_ | oh fine.. ya little bit | 12:02 |
| daemontool | git clone https://github.com/openstack/freezer | 12:02 |
| daemontool | cd freezer | 12:02 |
| jith_ | ok | 12:02 |
| daemontool | gedit README.rst | 12:03 |
| daemontool | edit | 12:03 |
| daemontool | git add README.rst | 12:03 |
| jith_ | will edit README file.. then how to send it?? | 12:03 |
| daemontool | git commit | 12:03 |
| jith_ | oh ok | 12:03 |
| daemontool | add to the commit message a description | 12:04 |
| daemontool | also add a bug previously opened on https://bugs.launchpad.net/freezer | 12:04 |
| daemontool | then | 12:04 |
| daemontool | git review -v | 12:04 |
| daemontool | and wait someone from us will review it | 12:04 |
| daemontool | so you have your first contribution in openstack :) | 12:04 |
| jith_ | done.. | 12:05 |
| *** EmilDi has joined #openstack-freezer | 12:05 | |
| *** reldan has joined #openstack-freezer | 12:05 | |
| jith_ | but couldnt get.. also add a bug previously opened on "" | 12:05 |
| jith_ | didnt get exactly | 12:05 |
| jith_ | sure :).. thanks for the great idea | 12:06 |
| jith_ | i didnt get this.. also add a bug previously opened on https://bugs.launchpad.net/freezer... | 12:07 |
| *** daemontool_ has joined #openstack-freezer | 12:07 | |
| daemontool_ | jith_, clone the repo | 12:07 |
| daemontool_ | and add to the readme the information | 12:07 |
| daemontool_ | that's the most important part | 12:07 |
| daemontool_ | :) | 12:07 |
| daemontool_ | after that is done | 12:08 |
| daemontool_ | let me know | 12:08 |
| daemontool_ | and we'll move forward | 12:08 |
| daemontool_ | :) | 12:08 |
| *** daemontool has quit IRC | 12:10 | |
| jith_ | daemontool_ : thanks.. sure sure | 12:13 |
| daemontool_ | :) | 12:18 |
| *** daemontool__ has joined #openstack-freezer | 12:34 | |
| *** daemontool_ has quit IRC | 12:37 | |
| *** EmilDi has quit IRC | 12:58 | |
| reldan | szaher , m3m0 , daemontool_ , vannif_ , slashme : Guys please tell me what we should do with this https://review.openstack.org/#/c/280602/ ? I can abandon it or should we accept it? Because from my point of view, this change is able to make our life much easier. | 13:22 |
| daemontool__ | I think we should to forward with that | 13:29 |
| daemontool__ | s/to/go/ | 13:29 |
| reldan | daemontool__: So, I’m going to work with comments then. | 13:34 |
| daemontool__ | reldan, yes please | 13:34 |
| m3m0 | let's move forward with this | 13:36 |
| daemontool__ | m3m0, ty | 13:37 |
| slashme | +1 | 13:37 |
| reldan | Thank you | 13:37 |
| reldan | Fix for metadata curr_level https://review.openstack.org/#/c/286592/ | 13:37 |
| reldan | Seems that our robot is dead | 13:37 |
| daemontool__ | m3m0, I think most of your comments make sense https://review.openstack.org/#/c/280602/ | 14:02 |
| daemontool__ | reldan, do you agree/disagree with comments from m3m0 szaher there? | 14:02 |
| openstackgerrit | Jin Nan Zhang proposed openstack/freezer-web-ui: Fixed importError urlresolvers https://review.openstack.org/286094 | 14:03 |
| reldan | daemontool_: do you like name mod for this? I have no strong opinion, if everyone agree to go with mods - let’s go | 14:04 |
| *** EmilDi has joined #openstack-freezer | 14:04 | |
| reldan | daemontool_: Here https://review.openstack.org/#/c/280602/4/freezer/hook/hook.py everything is good, but I suppose we should add it later | 14:05 |
| reldan | daemontool_: https://review.openstack.org/#/c/280602/4/freezer/hook/mongo.py yes, we should - but I suppose it should be done later. The idea of this pull request is structure refactoring. | 14:05 |
| reldan | https://review.openstack.org/#/c/280602/4/freezer/hook/mysql.py - the same | 14:06 |
| reldan | https://review.openstack.org/#/c/280602/4/freezer/job.py - really good idea, but should think about it a little bit more. Suppose it should be next step | 14:07 |
| reldan | https://review.openstack.org/#/c/280602/4/freezer/lvm.py - great idea, next step | 14:07 |
| zhangjn | who can review some patch for freezer-web-ui? | 14:07 |
| reldan | daemontool__: So let me change hook to mod and add todo’s with good comments from szaher and m3m0 | 14:08 |
| daemontool__ | m3m0, is this good to go? https://review.openstack.org/#/c/278366/1 | 14:10 |
| daemontool__ | did slashme tested in your env? | 14:10 |
| zhangjn | How to test the patch? | 14:13 |
| zhangjn | I can test it | 14:14 |
| zhangjn | https://review.openstack.org/#/c/278366/ | 14:15 |
| daemontool__ | m3m0, ^^ | 14:17 |
| daemontool__ | slashme, ^^ | 14:17 |
| daemontool__ | you have to enable TLS on the freezer-api | 14:18 |
| daemontool__ | then test if the web ui works | 14:18 |
| zhangjn | thanks, I will try it. | 14:19 |
| *** EmilDi has quit IRC | 14:28 | |
| m3m0 | TLS support in freezer-api is done with the config file, so no need to modify the code daemontool | 14:31 |
| m3m0 | yes it is good to go | 14:32 |
| m3m0 | zhangjn: which commit do you want review? | 14:33 |
| zhangjn | https://review.openstack.org/#/c/286094/ | 14:36 |
| m3m0 | there is one issue now, we should work with the stable/liberty version of horizon | 14:36 |
| daemontool__ | m3m0, so do we need to merge this code? https://review.openstack.org/#/c/278366/ | 14:37 |
| m3m0 | could you try freezer-web-ui with that version? as soon as we move to mitaka we should be able to merge this patch | 14:37 |
| zhangjn | This code looks good. | 14:37 |
| zhangjn | but not to test it. | 14:38 |
| m3m0 | agree with you, we need to start working on that | 14:38 |
| m3m0 | as soon as I finish the python-freezerclient I'll sync up with the ui for the testing | 14:39 |
| m3m0 | would you like to participate? | 14:39 |
| m3m0 | yes daemontool | 14:39 |
| m3m0 | good to go | 14:39 |
| daemontool__ | ok | 14:39 |
| *** EinstCrazy has joined #openstack-freezer | 14:39 | |
| m3m0 | zhangjn: do you know what's the best way to deploy javascrit code in horizon? | 14:40 |
| zhangjn | now stable/liberty freezer-web-ui is available? | 14:40 |
| m3m0 | yes an stable/liberty is available for freezer-web-ui | 14:42 |
| zhangjn | master is unavailable now? | 14:42 |
| zhangjn | ok, I will change the branch to stable/liberty to learn. | 14:43 |
| m3m0 | horizon stable/liberty -> freezer-web-ui stable/libery and master, horizon master -> we need to start working on that and your patch will help a lot | 14:43 |
| m3m0 | I can give you a code walktrhough if you like | 14:43 |
| zhangjn | thanks, I will speed some time to learn it and ask for you. | 14:46 |
| m3m0 | feel free to ask me anything | 14:47 |
| m3m0 | my email is guillermo.ramirez-garcia@hpe.com | 14:47 |
| daemontool__ | m3m0, are you available to give the code walkthrought next thursday? | 14:47 |
| daemontool__ | like we did last thursday | 14:47 |
| daemontool__ | this week vannif_ could give the it for the scheduler | 14:47 |
| daemontool__ | vannif_, ^^ | 14:47 |
| m3m0 | this week I can't I was planning everything for the 4th week | 14:48 |
| daemontool__ | yes | 14:49 |
| daemontool__ | this week we can do the scheduler | 14:49 |
| *** EinstCrazy has quit IRC | 14:50 | |
| *** EinstCrazy has joined #openstack-freezer | 14:50 | |
| zhangjn | m3m0: is friendly to me. LOL | 14:52 |
| *** daemontool_ has joined #openstack-freezer | 14:54 | |
| *** daemontool__ has quit IRC | 14:56 | |
| daemontool_ | ok flight booked, I'll be in Galway from 14-17 of March :) | 15:05 |
| slashme | daemontool_: Nice | 15:11 |
| *** EinstCrazy has quit IRC | 15:12 | |
| m3m0 | nice :) | 15:17 |
| reldan | daemontool_: Give me please +workflow https://review.openstack.org/#/c/284637/ | 15:24 |
| *** daemontool__ has joined #openstack-freezer | 15:24 | |
| reldan | daemontool__: Give me please +workflow https://review.openstack.org/#/c/284637/ | 15:24 |
| reldan | Guys give me +workflow please https://review.openstack.org/#/c/284637/ | 15:27 |
| *** daemontool_ has quit IRC | 15:28 | |
| dmellado | reldan: isn't the patch name quite misleading? | 15:44 |
| reldan | dmellado: Probably you are right | 15:44 |
| reldan | dmellado: But we have very strange approach for receiving arguments | 15:45 |
| openstackgerrit | Merged openstack/freezer-web-ui: fixed compilemessages error when install use devstack https://review.openstack.org/283631 | 15:45 |
| dmellado | what would be the usecase after the patch? I saw m3m0 asking in the PS but didn't see any reply | 15:45 |
| *** dschroeder has joined #openstack-freezer | 15:46 | |
| reldan | dmellado: everything should work perfectly. Currently we have a dict called DEFAULT_PARAMS and default params in cli arguments. | 15:47 |
| m3m0 | ok guys, this is the temp repo for python-freezerclient https://github.com/memogarcia/python-freezerclient/commits/master | 15:47 |
| reldan | and we can just remove duplicated default args from cli, because they just have no sense | 15:48 |
| dmellado | reldan: how about chaning the commit message to somthing like 'Move default params to dict' instead? | 15:48 |
| dmellado | s/chaning/changing | 15:48 |
| reldan | for example for confg file - we are still have defaults, but from dict | 15:48 |
| reldan | dmellado: They already there | 15:48 |
| reldan | I didn’t move, just delted it | 15:48 |
| * dmellado rechecking | 15:48 | |
| m3m0 | szaher https://github.com/memogarcia/python-freezerclient/commits/master | 15:49 |
| openstackgerrit | Merged openstack/freezer: Removing defaults values from cli. https://review.openstack.org/284637 | 15:49 |
| dmellado | reldan: yep, then 'Change **** values' ;) | 15:49 |
| dmellado | anyway xD | 15:50 |
| dmellado | got merged, but It'd be great to get a more accurate commit message for next times, otherwise it's hard to get the meaning | 15:50 |
| reldan | dmellad: Agree, but our code just sometimes have no meaning :) | 15:51 |
| reldan | dmellado: As far as I understand. If we have command like that freezerc —config /tmp/config —storage=local —container=/tmp/backup | 15:52 |
| reldan | dmellado: we want to read config and then overright storage and container | 15:52 |
| dmellado | override? | 15:53 |
| dmellado | cool | 15:53 |
| reldan | dmellado: Yes. But here is a problem. We have default values. And let’s say I have —storage=local in config and my command is freezerc —config /tmp/config —storage=swift | 15:54 |
| reldan | dmellado: | 15:54 |
| reldan | cli_options = dict([(x, y) for x, y in CONF.iteritems() if y != | 15:54 |
| reldan | defaults.get(x) and y is not 'False']) | 15:54 |
| dmellado | I see | 15:55 |
| reldan | So here, I would like to override not default value from config with default value swift | 15:55 |
| reldan | and that code will check that swift is default and don’t override it | 15:55 |
| reldan | cli_options = dict([(x, y) for x, y in CONF.iteritems() if y is not None]) | 15:55 |
| reldan | This code should work, because I have None for every default parameter from cli | 15:56 |
| reldan | and there is no possiblity to provide None as argument, that I may be sure - if I have None, user didn’t provide any value | 15:56 |
| dmellado | reldan: got it, thanks for the explanation ;) | 16:00 |
| reldan | dmellado: Thank you. I will try to have better description for my commits :) | 16:01 |
| dmellado | thanks! ;) | 16:01 |
| m3m0 | daemontool__ can you create the pypi project for python-freezerclient ? | 16:01 |
| openstackgerrit | Fabrizio Vanni proposed openstack/freezer: lvm snapshot unique name https://review.openstack.org/286705 | 16:43 |
| openstackgerrit | Eldar Nugaev proposed openstack/freezer: Refactoring of mysql, msserver and mongo code. https://review.openstack.org/280602 | 16:46 |
| openstackgerrit | Fabrizio Vanni proposed openstack/freezer: lvm snapshot unique name https://review.openstack.org/286705 | 16:53 |
| m3m0 | . | 17:08 |
| *** reldan has quit IRC | 18:06 | |
| *** openstackgerrit has quit IRC | 18:18 | |
| *** openstackgerrit has joined #openstack-freezer | 18:18 | |
| daemontool__ | m3m0, yes, but does it works? | 18:42 |
| daemontool__ | I mean, if we remove the apiclient and use the python-freezerclient as dependancy in the web ui and scheduler | 18:43 |
| daemontool__ | does it works? | 18:43 |
| *** dmellado has quit IRC | 18:48 | |
| *** openstack has joined #openstack-freezer | 19:17 | |
| m3m0 | yes :) but not all the functionality is there yet | 19:20 |
| *** openstackgerrit has joined #openstack-freezer | 19:20 | |
| m3m0 | daemontool__ | 19:20 |
| daemontool_ | m3m0, ok | 19:21 |
| *** smekel_ has quit IRC | 19:25 | |
| *** smekel_ has joined #openstack-freezer | 19:31 | |
| *** daemontool__ has joined #openstack-freezer | 20:05 | |
| *** daemontool_ has quit IRC | 20:09 | |
| *** reldan has joined #openstack-freezer | 20:31 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!