Devart Forums

I'll give the new beta a try and let you know!

Finally, someone else has the same problem!

Still haven't found a solution, and the only answer I've received from Devart was to use their connection pooling builtin, which isn't helpful as there are multiple apps across different programming languages and platforms that all need to access the database.

Unfortunately I've just resorted to connecting directly to the database server with dotConnect, and bypassing pgpool, which is a very fragile and annoying workaround. Everytime a new version comes out I always hope it's fixed, but no joy thus far. Every other client library I've tried works fine with pgpool, except for this one.

If you find a solution, please let me know...

Mike

Hello all,

I am using Entity Developer to create an Entity Framework model from an existing PostgreSQL database. One of the columns is of datatype 'xml', which I believe is an official, out-of-the-box, supported data type as of PostgreSQL 8.3, if not sooner.
When attempting to add the table in question to the model, I get the following error:

"The column 'column1' has unsupported data type 'xml'. Cannot create class for entity 'tablename'."

Am I doing something wrong, or is there an extra step I need to go through or something?

I'm using the latest .39 version, PostgreSQL v8.3.

Any thoughts as to why this would be a problem, any other possible workarounds other than disabling dotConnect connection pooling (which has no effect), or anything else to try to help narrow down the problem?

It's not that I don't want to use the dotConnect connection pooling, it's that we have multiple applications, written in C#, Python, PHP, etc.. running on multiple platforms, so to enhance database performance across the board we run connection pooling on the server.
The first thing I tried was turning off dotConnect connection pooling but it had no effect whatsoever. Any other changes that might help?

Looking at it from the server side, the connection is made, and I can see the query running from pgadmin, but it continues running and never returns.

Has anyone ran into any problems, or know of any bugs that cause problems when running pgpool on the database server?
We are using pgpool-II for connection pooling, and no other applications have a problem, but for some reason whenever I attempt to use dotConnect the transaction never completes. Looking on the database server, I see the connection as active, but regardless of the query it never returns.
If I connect directly to the port the database is listening on (pgpool is running on port 5432), with the exact same connection information and query, everything works great, but connecting to pgpool isn't working.
Anyone?

In case anyone else runs into this problem, the new dotConnect 4.0 release maps my database just fine in just a few minutes! Problem solved!

Hello all,

I have a current PostgreSQLDirect.Net subscription from 3.75, and I'm trying to find the licensed version of dotConnect for PostgreSQL to download. It's not listed on the page to download the licensed PostgreSQLDirect.NET versions, so any help would be appreciated...

Mike

I am evaluating PostgreSQLDirect.NET and am having problems generating the initial edmx model. The problem is the process hangs for at least 24 hours, whether I use EdmGen(2) from the command line, or use the Visual Studio tool.

On the server side, the PgAdmin logs show a query that looks like this being executed:

Select "Project6"."C11" AS "C1", "Projects6"."CatalogName", "Projects6"."SchemaName", etc... being executed.

Going with the theory that the database was overloaded, or the amount of data in the database was affecting the efficiency, I exported the schema only using pg_dump, and loaded it into a freshly installed copy of PostgreSql 8.3 on my local machine. However, even with an empty, unused database, the process was still taking just as long, with no results.

As I've read there are issues with how efficient the entity framework's mapping tools (EdmGen, EdmGen2) are when mapping large numbers of tables, I attempted to map just a single table, with the exact same results; no results written, no progress made whatsoever.

There is no way I can provide the schema to support or publicly, but given that I have the schema, is there anyway to generate the edmx directly from the schema definition file, without interrogating the database and reverse engineering it?

Does anyone have any insight or ideas as to how to get around this? Any help would be greatly appreciated.