[openstack-dev] Program Proposal: Trove

Michael Basnight mbasnight at gmail.com
Wed Jul 17 02:58:38 UTC 2013


On Jul 16, 2013, at 7:45 PM, Haomai Wang wrote:
> 
> 在 2013-7-17,上午4:42,Michael Basnight <mbasnight at gmail.com> 写道:
> 
>> On Jul 16, 2013, at 12:49 PM, Mark McLoughlin wrote:
>> 
>>> Hey,
>>> 
>>> On Tue, 2013-07-16 at 10:37 -0700, Michael Basnight wrote:
>>>> Official Title: OpenStack Database as a Service
>>>> Initial PTL: Michael Basnight <mbasnight at gmail.com>
>>>> 
>>>> Mission Statement: To provide scalable and reliable Cloud Database as
>>>> a Service functionality for both relational and non-relational
>>>> database engines, and to continue to improve its fully-featured and
>>>> extensible open source framework.
>>> 
>>> Seems fine to me, but I'd see adding non-relational support as an
>>> expansion of Trove's scope as approved by the TC.
>>> 
>>> I know we discussed whether it should be in scope from the beginning,
>>> but I thought we didn't want rule out the possibility of an entirely new
>>> team of folks coming up with a NoSQL as a Service project.
>> 
>> Cant disagree with this, because of the initial TCT ruling. FWIW we are working on some NoSQL stuff _in_ trove at present. Maybe i should bring it up at the next TCT meeting? Ive done a redis POC in the past and can show the code for that. It was before the rename and has a small ammt of bitrot but its something i can definitely show to the group.
> +1, I'm interested in Trove and I want to do some works for it. NoSQL is easier to control than SQL database and I want to join it to implement LevelDB.

Great! Feel free to join in. The API is extensible enough at present to allow you to implement LevelDB, or any other NoSQL data store. Its _almost_ as simple as adding a custom guest manager [1], and making sure you edit the api config to specify service_type=name_of_service_type. There are a few more small gotchas, but thats a good 80% of the implementation. You can also create strategies to define how you backup/restore 'service_type', but i wont get into that on list. We are also working on a cluster api to allow cluster instrumentation, which i will be sending out in the next few weeks for the community to scrutinize.

Find me on irc, #openstack-trove, user hub_cap for more information. 

[1] https://github.com/openstack/trove/tree/master/trove/guestagent/manager


More information about the OpenStack-dev mailing list