How can i add new data source type other than the currently listed source types.
Is it possible to add other relational database source to dremio.
Currently in dremio it's connecting hive through metastore is it possible to connect through jdbc.
Thanks & Regards
Rahul
Your best bet is the ARP format, which provides a generic interface for all JDBC libraries. You can see the docs, and an implementation used for SQL Lite here.
https://github.com/dremio-hub/dremio-sqllite-connector#arp-file-format
Related
good day. i have some theoretical question.
i made application that uses neo4j to traverse graphs and make some external calculations on each node. traverse process initiated by external event.
during traverse process nodes properties may change according calculation results.
this made by
new GraphDatabaseFactory().newEmbeddedDatabase(new File (this.DB_PATH));
GraphDatabaseService.traversalDescription();
now i'd like to use neo4j OGM to use domain classes in path extenders code instead self-made node-wrappers. also i`d like to use domain classes to populate DB with nodes and relations , and change properties of nodes from external source (REST for example).
But i can't find way to get Session or SessionFactory to existing and operational database.
is it possible to connect to database , created as embedded and conducting some calculations in one module , from another module by BOLT protocol, execute some queries, use OGM features?
The OGM cannot be used from path expanders / traversals (at least not easily).
To use OGM with embedded database you have 2 options
use the standard configuration as described in the reference documentation
set use Components.setDriver to set the driver manually, see e.g. this answer for details
Im new to breeze js, I understand that breeze has it's own query language related to OData
But I am trying to get breeze js working with a custom query language, as an example FIQL is
one of the form to make queries to backend, can we get breeze run with this type of query language.
Overview of FIQL :-
http://jaxenter.com/tutorial-smarter-search-with-fiql-and-apache-cxf-46876.html
(this is not a technical answer, at the very most it's an advice)
As mentionned from their website documentation :
Today, out of the box, the Breeze product ships with adapters for the
ASP.NET Web API and for OData. It also ships with .NET components that
interface with the Entity Framework and that generate Breeze metadata
from an Entity Framework model; that EF model could be developed code
first or database first. The Breeze client is in no way limited to
these technologies; they are merely the first backend components
available; we’d be thrilled to help you adapt Breeze to your preferred
server stack.
First question is whether
1) you want breeze to actually create FICL queries on the client and send them to a FICL service. This involves translating a client side breeze EntityQuery into FICL url syntax. If so you would need to create a new a 'uriBuilder' adapter.
2) you want to translate a breeze query already sent to your service into FIQL syntax on the server so that you can further refine the query and then send it on to your FICL service. If so you would need to create a breeze DataService endpoint.
Both of these are certainly possible.
If you are looking at the 1st option, take a look at the 'b000_uriBuilder.xxx' files in the breeze GitHub repo under the 'src' directory. There should be two 'uriBuilder.xxx' files, one which translates an EntityQuery into OData syntax and another that translates the EntityQuery into a new 'json' query syntax. There is also additional information on this 2nd syntax here:
http://www.getbreezenow.com/documentation/query-using-json
Both of these 'uriBuilder's make use an internal 'visitor' framework that we have not yet documented but should be reasonably understandable by looking at the source.
If you are looking at the 2nd option, I would recommend looking at the link mentioned above and then taking a look at the 'breeze.node.js' repo and the 'breeze-sequelize' subdir within it. In this repo we take a serialized server side breeze EntityQuery and translate it into a 'Node' Sequelize query that we can apply to a variety of backends.
Another option is to contact our professional services arm, at breeze#ideablade.com and ask them to assist in building the adapters for you or to ask for more detailed technical help.
Does anyone know if it is possible to use the Entity Framework Code First approach when developing applications using Monotouch? I appear to be at a stumbling block with the lack of System.Data.Common.DbProviderFactories support in the System.Data library.
If this is not possible, what other similar options are available?
I'd look at SQLiteNet, link to the Component Store : Sqllite Net
That will help you get started. I am assuming, your post relates to using a local SQLLite database to store or cache data on the client mobile application.
Basically, you create your own classes, with attribute markups. You can delete and Insert, but updates to your objects require sql statements. There is a link at the top of that getting started page, to the full documentation.
Currently I'm developing a webapp which uses Oracle as a database. I've already been informed that my webapp has to work also on Sybase ASE. My Oracle database uses sequences to generate IDs and I mapped my domain classes to use those sequences. As far I know Sybase ASE doesn't have sequences but identities. And here is my problem. What to do with code/configuration to work with Oracle and Sybase. Some "conditional mapping" of ID for each domain would do.
I just don't want to comment/uncomment lines in mappings every time when I'm building webapp for other database than already mapped. Have anybody any idea what to do?
There is another danger: another differences between Oracle and Sybase which have influence on mapping (I don't generate database structure using dbCreate = "create") . At this point I don't see anything. Any experiences?
EDIT:
It turned out that it's not possible make one mapping for both databases. My Oracle structure uses quotes in column and table names. And quotes are not allowed in Sybase ASE. My (not nice and pretty) solution is changing (comment/uncomment) mapping when building webapp on some particular platform.
Btw I'm huge fan of Grails and this mapping thing seems to be a drawback. If I was using "pure" Hibernate I'd have two mappings in xml files and I'd change them depending on an underlying database. But Grails doesn't give me possibility to have two mappings.
I don't know if grails enable to you to configure coditional mapping for id generation but I think that maybe you don't need use neither sequences or identities, and use other hibernate id generator strategy: http://grails.org/doc/1.3.7/ref/Database%20Mapping/id.html.
EDIT:
Reviewing the hibernate documentation I found that is possible use sequences or identites according to capacities of database, from http://docs.jboss.org/hibernate/orm/3.3/reference/en/html/mapping.html#mapping-declaration-id:
All generators implement the interface
org.hibernate.id.IdentifierGenerator. This is a very simple interface.
Some applications can choose to provide their own specialized
implementations, however, Hibernate provides a range of built-in
implementations. The shortcut names for the built-in generators are as
follows:
...
native:
selects identity, sequence or hilo depending upon the capabilities of
the underlying database.
So you maybe will need a configuration like this:
static mapping = {
id generator:'native'
}
On one server there are more than 20 databases with identical structure but different data. I need to collect some of the data (the same queries) from all databases and store in new database which is located on another server. I decided to use ASP.NET MVC 2 but it doesn't seem logical to use more than 20 "LINQ to SQL Classes" (.dbml) files because the structure is the same for all databases and it's repeating if I use so many of these files. Is there a simple way to use one .dbml file (for remote databases) but change only connection string?
I agree that you really wouldn't want to use MVC as that is a web framework and has nothing to do with moving data around.
You can also look into using an ETL tool to accomplish this task. I have used RhinoETL in the past successfully to accomplish something similar.
There are also multiple posts on this site discussing ETL tools. For example, check the following link - https://stackoverflow.com/questions/51198/what-etl-tool-do-you-use
According to this, you can pass in a connection string with the dataContext constructor. So theoretically, you should be able to have one dbml file, but you can instantiate multiple instances of your data context, each with a different database connection string specified. Each context should then point to their respective database and allow you to work with multiple databases.
Why do you want to use ASP.NET MVC at all? ASP.NET is for web UI, not data warehousing (except when you need to display cubes). Looks like you use SQL Server. If that is true you can utilize Integration Services (ex DTS) to do the job.