| *** rcernin has joined #openstack-trove | 00:00 | |
| *** sapd1 has quit IRC | 00:34 | |
| *** rcernin has quit IRC | 03:06 | |
| *** rcernin has joined #openstack-trove | 03:11 | |
| __ministry | Hi, lxkong | 03:18 |
|---|---|---|
| __ministry | I was contributing to trove for support meta_data field. but I see we can crud with metadata in link: https://docs.openstack.org/python-troveclient/ussuri/cli/trove.html#trove-metadata-create | 03:18 |
| __ministry | Can you tell me about this metadata? | 03:18 |
| __ministry | I also see in python-trove client have class: https://github.com/openstack/python-troveclient/blob/master/troveclient/v1/metadata.py but I can't run it. | 03:20 |
| *** sapd1 has joined #openstack-trove | 03:23 | |
| *** sapd1 has quit IRC | 05:09 | |
| *** yasemind has joined #openstack-trove | 05:26 | |
| *** Reepicheep has quit IRC | 05:52 | |
| *** Reepicheep has joined #openstack-trove | 05:53 | |
| yasemind | lxkong lxkong in stable/victoria devstack, postgre is working successfully, it is suprising for me but i didnt understand why not running in our HA stable/victoria test enviroment, building with openstack-ansible, do you have any idea? | 06:30 |
| lxkong | > <__ministry> I was contributing to trove for support meta_data field | 06:34 |
| lxkong | i am still not sure about your requirements | 06:35 |
| lxkong | yasemind: i've no idea. If it's working in devstack, it should also work in your deployment. I don't know how openstack-ansible deploys trove. | 06:35 |
| lxkong | i remember in Victoria, there are some functions missing for postgresql | 06:36 |
| lxkong | I've proposed several patches in Wallaby for postgresql, it's better than V. | 06:36 |
| yasemind | do you recommend to use W ? | 06:37 |
| lxkong | yasemind: depends on your requirements | 06:37 |
| yasemind | we want to support mysql, mariadb and postgresql | 06:38 |
| yasemind | after, redis cassansdra etc. | 06:38 |
| yasemind | mongo | 06:38 |
| lxkong | yasemind: you can try first in devstack to see if the existing functions could satisfy your requriements. | 06:39 |
| __ministry | lxkong: I have 3 type of database instance that are: basic, premium, enterprise, each type will be limited data transfer, bandwidth, iops, ... I want save it to metadata to not need an other database to save this informations. Can you give advice to me? | 06:39 |
| lxkong | if not, we could see if the missing features are supported in W, if not neither, we could discuss how we could implement | 06:40 |
| yasemind | lxkong okey i will build to devstack with W | 06:40 |
| lxkong | __ministry: if tags are supported, do you think it will do what you want? | 06:40 |
| __ministry | Yep, sure. | 06:42 |
| __ministry | but can you talk about metadata in this link: https://docs.openstack.org/python-troveclient/ussuri/cli/trove.html#trove-metadata-create? I can't use it. | 06:43 |
| lxkong | __ministry: the links you provided make sense. I remember Trove supported metadata before, but I need to check if it's still the case now. | 06:43 |
| __ministry | yes, many thanks. | 06:44 |
| lxkong | When I started contributing Trove, I didn't see the metadata thing in API doc, not in trove openstack CLI implementation. | 06:45 |
| *** ahsen has joined #openstack-trove | 06:50 | |
| __ministry | Can you check it? I thinks this is an important field, it can help we very other if you want save some data without other database. | 06:51 |
| lxkong | __ministry: I just checked, I didn't see there is any db table related to metadata | 06:53 |
| lxkong | so to make consistent with other openstack services, we could add tag support | 06:54 |
| lxkong | __ministry: are familar with tag support in other openstack services? | 06:54 |
| lxkong | are you familiar with tag support in other openstack services? | 06:54 |
| lxkong | you can take a look at their API doc first, have a try in CLI, then (if you want to save time) read their source code for implementation. | 06:57 |
| lxkong | tag support shouldn't be related to datastore functions, so the implementation should only touch API layer. | 06:58 |
| __ministry | i can familiar with metadata in some openstack service, and tag - i just see in neutron to set for networks and subnets. | 07:04 |
| __ministry | but, it is array not string. | 07:05 |
| lxkong | __ministry: Nova: https://docs.openstack.org/api-ref/compute/#server-tags-servers-tags; Octavia: https://docs.openstack.org/api-ref/load-balancer/v2/index.html#filtering-by-tags | 07:07 |
| lxkong | tag is an array of strings. | 07:07 |
| lxkong | __ministry: yeah, you can choose to implement metadata (key-value pairs) | 07:08 |
| lxkong | up to you. | 07:08 |
| __ministry | by the way, can you tell me why you suggest me use tag? | 07:08 |
| lxkong | in the history, tags were recommended in openstack services for filtering | 07:10 |
| lxkong | but that's history, not many services were following that way | 07:10 |
| __ministry | yep, I want use metadata to save custom my information like as other services in openstack. because not need to filtering. | 07:15 |
| __ministry | I just want read data from that. | 07:15 |
| __ministry | so can you approved my commit? | 07:19 |
| *** e0ne has joined #openstack-trove | 07:22 | |
| *** rcernin has quit IRC | 07:29 | |
| *** sapd1 has joined #openstack-trove | 08:15 | |
| *** tosky has joined #openstack-trove | 08:28 | |
| *** rcernin has joined #openstack-trove | 08:48 | |
| yasemind | lxkong i want to functional test in our test environment, how can i run functional test without devstack? | 08:56 |
| *** rcernin has quit IRC | 09:00 | |
| *** rcernin has joined #openstack-trove | 09:52 | |
| *** rcernin has quit IRC | 10:07 | |
| *** rcernin has joined #openstack-trove | 10:07 | |
| *** rcernin has quit IRC | 10:20 | |
| *** rcernin has joined #openstack-trove | 10:27 | |
| *** rcernin has quit IRC | 10:38 | |
| lxkong | yasemind: you can run trove tempest | 10:43 |
| lxkong | __ministry: i will review your code before approving | 10:43 |
| __ministry | yep, thank you. | 10:43 |
| yasemind | lxkong do trove tempest run in stable/victoria ? | 10:44 |
| *** e0ne has quit IRC | 10:51 | |
| *** rcernin has joined #openstack-trove | 10:52 | |
| *** e0ne has joined #openstack-trove | 10:56 | |
| *** e0ne has quit IRC | 10:58 | |
| lxkong | __ministry: I left a comment in https://review.opendev.org/c/openstack/trove/+/782829 | 10:59 |
| lxkong | Adding a field in instance table will make it hard for filtering by metadata | 11:00 |
| lxkong | filtering by metadata is a very useful feature for users. | 11:00 |
| lxkong | and please use `metadata` instead of `meta_data` | 11:01 |
| *** rcernin has quit IRC | 11:05 | |
| *** e0ne has joined #openstack-trove | 11:47 | |
| *** e0ne has quit IRC | 12:09 | |
| __ministry | yes, and I also want add metadata to nova vm. so I think we still need pass to task manager. | 12:13 |
| *** sapd1 has quit IRC | 12:25 | |
| *** sapd1 has joined #openstack-trove | 13:40 | |
| *** rcernin has joined #openstack-trove | 14:01 | |
| *** rcernin has quit IRC | 14:06 | |
| *** __ministry1 has joined #openstack-trove | 15:09 | |
| *** ahsen has quit IRC | 15:39 | |
| *** __ministry1 has quit IRC | 16:53 | |
| *** yasemind has quit IRC | 17:03 | |
| lxkong | > __ministry> yes, and I also want add metadata to nova vm. so I think we still need pass to task manager. | 19:31 |
| lxkong | ok, make sense | 19:31 |
| *** e0ne has joined #openstack-trove | 20:51 | |
| *** e0ne has quit IRC | 22:13 | |
| *** e0ne has joined #openstack-trove | 22:18 | |
| *** e0ne has quit IRC | 22:19 | |
| *** rcernin has joined #openstack-trove | 22:24 | |
| *** rcernin has quit IRC | 22:24 | |
| *** rcernin has joined #openstack-trove | 22:24 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!