Thank you both. You have confirmed what I needed to dig in. I'm excited to be working with orientdb.
On Tuesday, April 21, 2015 at 3:54:37 PM UTC-5, Colin wrote: > > The embedded OrientDB server may be used in distributed mode just fine, > even with other non-embedded instances. > > -Colin > > On Tuesday, April 21, 2015 at 2:31:02 PM UTC-5, [email protected] > wrote: >> >> Thank you Luigi. That is what my research suggested, I just wanted to >> make sure I was on the right track before digging into code and design >> testing. >> >> It is then possible to scale the embedded database across multiple >> servers, yes? >> >> On Tuesday, April 21, 2015 at 9:11:20 AM UTC-5, Luigi Dell'Aquila wrote: >>> >>> Hi Chris, >>> >>> I do not have detailed data about differences in performance between >>> Python and Java interfaces, anyway Java API in embedded server is the most >>> performing in absolute, you can expect more than twice as the performance >>> IMHO. >>> >>> Embedding OrientDB in Java applications is a very common practice, your >>> application will be able to expose REST services on its own port, then you >>> can decide to just use OrientDB as a plocal db (without binary/REST >>> interfaces enabled) or to start it as a full featured embedded server (with >>> plugins, binary and rest interfaces). There is no need to touch the source >>> code, basically what you need to do is just open a db connection with >>> "plocal:" path. >>> >>> Luigi >>> >>> >>> 2015-04-21 15:46 GMT+02:00 <[email protected]>: >>> >>>> I am creating a restful server that will more-less expose premade types >>>> of queries. For instance: GET /relationships/mutual-friends?a=11&b=14. >>>> That >>>> is an incredibly simple example. The end server is providing relational >>>> data for deep learning models. The server itself needs to be able to >>>> handle >>>> oAuth, user access lists, and these premade queries. That's all. >>>> >>>> Performance is absolutely crucial as is the future ability to replicate >>>> and scale the database. Most of the rest of the stack is written in python >>>> with client-side interfaces in HTML5/CSS/JS so my first reaction is to use >>>> a python microframework for the rest server and communicate through binary >>>> connections. But,* is there a significant performance gain between >>>> this an using a native Java Api on an embedded server?* Has anyone >>>> done benchmarks? >>>> >>>> Assuming that it is close to twice as performant to use the native api >>>> on an embedded database, my first thought would be to embed an orientdb >>>> into something like the SpringFramework. However, orientdb obviously has a >>>> restful architecture.* Is it possible/safe to extend this server >>>> architecture* to add new routes while keeping administrative functions >>>> locked away from the end-user? I do not want to modify the source, but >>>> wrap >>>> or extend it. >>>> >>>> Has anyone face a similar situation and has some feedback. It would be >>>> greatly appreciated. >>>> >>>> -- >>>> >>>> --- >>>> You received this message because you are subscribed to the Google >>>> Groups "OrientDB" group. >>>> To unsubscribe from this group and stop receiving emails from it, send >>>> an email to [email protected]. >>>> For more options, visit https://groups.google.com/d/optout. >>>> >>> >>> -- --- You received this message because you are subscribed to the Google Groups "OrientDB" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
