Showing posts with label requirement. Show all posts
Showing posts with label requirement. Show all posts

Monday, March 26, 2012

Offline Report Generation

Hi,

I have a requirement of generating very long reports (1-2hrs approx). Obviously the user would not like to wait for the report to come up to view.

Can I generate reports offline using SSRS 2005? If yes, then how can i submit the report for generation and re-direct the user to any other page, where he/she can see the status of the report (or continue to work on other pages).
If yes, then can I run an exe in the background to check the status of the report at regular intervals and notify the user once the report is completed? I checked the status of the report in the ExecutionLog table should be 'rsSuccessful', indicating the successful generation of the report.

If yes, then where (path) the reports generated offline are stored?

Thanks in advance.One option would be to schedule the reports and then deliver them to an email address or fileshare.
Another option would be to use the snapshot feature.
From BOL:

Report history is a collection of report snapshots. You can maintain report history by adding and deleting snapshots, or by modifying properties that affect report history storage. You can create report history manually or on a schedule.

To create report history, your role assignment must include the "Manage report history" task. To view report history, your role assignment must include the "View reports" task. Report history is available to all users who have access to the report. You cannot selectively enable or disable report history for a subset of users.

Snapshots in report history are identified by the date and time they were created. The date and time is based on when the query executed.

Monday, March 12, 2012

ODBC Virtual Driver

Hi,
I have the following requirement :

I need to have a way to intercept the SQL queries from an application written in VB and using a ODBC driver and modify the SQL queries before it goes through the ODBC driver and then to the database. This I need to do without modifying the original application.

The solution I have in mind is to write a ODBC virtual driver and configure my application to use my virtual ODBC driver. The ODBC virtual driver in turn will use the actual ODBC driver to the database. The virtual driver will basically intercept the SQL queries, modify it and then give it to the real ODBC driver.

My question is

1. Is this a feasible solution?
2. What should I do in order to implement the vitrtual ODBC driver.
3. Any pointer will be appreciated.

Thanks
Jake.NEver heard of virtual odbc driver, but you may find some information from http://www.microsoft.com site.|||Hi,
What I meant by Virtual driver is basically a ODBC proxy which can sit between my application and the ODBC driver and intercept the SQL queries.

Thanks
Jake

Originally posted by Satya
NEver heard of virtual odbc driver, but you may find some information from http://www.microsoft.com site.|||I'm working from memory here, but I think an ODBC Proxy something like you are describing is included as a VC project in the ODBC Driver SDK.

-PatP

Monday, February 20, 2012

ODBC Inserts are very slow

Hi,

I am new to the windows world. We use Informatica on UNIX for ETL process. We have a requirement to load approx. 200,000 rows to a MS SQL Server table . The table is not that big and it is a heap table (no indexes). Inserts are taking 69 rows/per minute. We are using DataDirect Closed 4.10 SQL Server ODBC driver.

SQL Profiler tells us that is is doing a row by row processing and using sp_execute procedure.

Is there a way we can speed up the ODBC process?

-Thanks in advance
srv

SQL Server Version:
Microsoft SQL Server 2000 - 8.00.818 (Intel X86) May 31 2003 16:08:15 Copyright (c) 1988-2003 Microsoft Corporation Standard Edition on Windows NT 5.0 (Build 2195: Service Pack 4)That would appear to be an older version of Informatica (pre-2003 installation). There was a problem that was fixed late in the year (2003-09 I think).

As a work-around, have Informatica generate a flat text file (tab delimited is my preference, columnar is often simplest, CSV is tolerable). FTP that file to your NT machine, then use DTS or BCP to load the file into SQL Server. You should have no trouble getting 10000 rows per minute from a mid-range workstation, and could get 1000000 rows per minute on a sufficiently beefy server.

-PatP|||If possible use DTS in SQL server which used BULK LOAD process for such load and for further information refer to the books online or http://www.sqldts.com website.
HTH|||We performed a test with Informatica 7.x and same odbc drivers and it 4 times faster. --Thanks