Jira issue originally created by user pmithrandir:
I'm looking for a scalable SQL database, and I found nuoDB(old nimbus DB) that seems to do the job on paper.
My company tend to encourage us to move to many small VM architecture in place of large machines.
I wanted to know if such a DB engine was going to be supported in the near future, or if it was now.
For what I saw, the only really scalable solution I have now is nosql DB as mongo DB with doctrine.
As Doctrine is a requirment on our application, I was condering if you had any feedback for us on this subject.
Best regards, and thank you for your work.
Comment created by stof:
The best way to have NuoDB support in DBAL is probably to contribute it.
Introducing new drivers is best done by people using them (because they actually need it, and because they can more easily test them)
You will need to provide 3 classes (and their tests):
Comment created by pmithrandir:
I already asked the question to NuoDB company, waiting their answer.
I'm absolutly not familiar with it, so I'm not very confident in the fact I would be able to provide a correct interface for this server.
I read mysql driver, shema and platform as an example. Driver and shema development should be dueable... but platform is totally out of my scope / scope... It would require a good knowledge of the database, and honnestly, I need only to perform 3 operations on it : INSERT / UPDATE and SELECT... without even a join. Would it be possible to let the other method empty and to let the community fill the blank ?
unfortunately, you need more than this in case you are using the ORM, because the ORM needs more methods from the platform (many of them has been added based on the needs of the ORM). However, the nuodb doc should only implementing these methods: doc.nuodb.com/display/doc/SQL+Reference
Thank you for your answer...
If we want to be compliant to Doctrine, is there a number of method that are required, and some other that are usefull?
I mean I can implement the method, but do nothing in some of them. The next user who need more will develop the module, or we will do later if needed.
The idea for me is to lower the cost, and to not pay on my project budget the entire connectivity...
I'm not againts speding 3-5 man day in this subject, but more will be an issue regarding the total amount of day I have on the project.