Help Needed -- Dynamically Loading Different Files/types
Nov 29, 2007
I am pretty new to SQL Server 2005 and SSIS. I am trying to develop a package that will dynamically load files into SS2005 based on the contents of a configuration table. The configuration table (see below for example) contains the path to the file, a flag indicating whether or not to process the flag, the type of file (specifies the nature of data -- financial, order, etc.) and some parameters specific to each file.
FileName
ProcessFlag
Type
ExcelTab
Param1
C:File1.xls
TRUE
1
Sheet1$
C:File2.xls
TRUE
1
Sheet1$
C:File3.xls
TRUE
1
Sheet1$
C:File4.xls
TRUE
2
Sheet1$
C:File5.xls
FALSE
2
Sheet1$
C:File6.csv
TRUE
3
C:File7.txt
TRUE
4
Right now I basically have a seperate sequence for each of the file types. The task in each sequence are virtually identical with the exception of the the data flow source in the data flow task (since the source file could be .xls, .txt, .csv). The first sequence ran fine in isolation, but when I linked a second sequence I started getting a Package Validation Error:
DTS_E_CANNOTACQUIRECONNECTIONFROMCONNECTIONMANAGER
I use the same variables (FileName, ExcelTab) for each of the sequences, so I am not sure if that causes te error. I also tried changing the ValidateExternalMetadata setting to false, since the connection variable wont be applicable until that sequence is being processed. I am not sure where to go here ... should I rearchitect how the package all together?
Is there a better/most efficient way to to architect this package to handle the different file types (develop a package for each type which is called by a master package, create one package with a different sequence container for each type, etc).
I am trying to do here is to load different flat files to different tables: For example, if the file name starts with "enrollment", then it goes to table "enrollment" table; if the file name starts with "student", then it goes to "student" table.
For now, I created a foreach loop container for the each different files. So it ended up using several foreach loop containers. I am wondering if there is a way just to use one foreach loop containters to process the loading.
Anyone shed some light on this?? Thank you very much for your help!
I am pretty new to SQL Server 2005 and SSIS. I am trying to dynamically load files into SQL Server 2005 using files/paths contained in tables. I have a key table in my SSIS package that defines which files should be processed, some default values that will be associated with each file (ie CompanyKey and PeriodKey) and the file path (see below for example):
CompanyKey
PeriodKey
ProcessFlag
FileName
2
61
FALSE
2005TB200501.xls
2
62
FALSE
2005TB200502.xls
2
63
FALSE
2005TB200503.xls
2
64
TRUE
2005TB200504.xls
2
65
TRUE
2005TB200505.xls
2
66
TRUE
2005TB200506.xls
2
67
TRUE
2005TB200507.xls
2
68
TRUE
2005TB200508.xls
2
69
TRUE
2005TB200509.xls
2
70
TRUE
2005TB200510.xls
2
71
TRUE
2005TB200511.xls
2
72
TRUE
2005TB200512.xls
The package I am developing is supposed to loop through this table checking for files with ProcessFlag = TRUE. For those files, it will load all of the records/columns (prefaced with the company and period keys) into a common table.
Do I have to manually create data sources for each file or is there a way to dynamically define the connection and process the connection?
Any assitance you could provide would be greatly appreciated!
I have a table structure like this EmpID LastName FirstName Emp_Picture 100| x |T |<BINARY> 200| W |W |<BINARY> .. .. ETC This table has 935 rows in it with Emp_Picture Blank.
How to insert the jpeg files into the Emp_Picture Column?? Do we have run the update statement for each and every employee or is there a way to get around this problem..
Is it possible to take a text file that contains multiple record types through the Data Transformation Service in MS SQL 7.0 and load each different record type into a seperate table?
I am new to sql server. I have some tiff files to load into sql database. The Server is 2005.
Can i do this without using any application like asp.net/c#. Is there any way to upload tiff files into tables using sql.The size of each image is approx 200-300kb.
I have tables with member information. The tiff file name is same as the member id. So i have to uplaod the image to the column in the member table with same id.
Can you guys please help me with this or suggest some articles/urls which use sql to upload tiff files.
I am writing a SSIS package to load a lot of Excel files. I use SQL statement to select the Excel data. However, I found it's hard to dynamical set the table name (Excel Tab name) - the user name the Tab differently.
In MySQL, I use "LOAD DATA INFILE 'my_path/data_file.txt'" to load datafrom a plain text file. Of course, the actual statement is a bit morecomplex once one considers the various options (e.g. comma delimited vstab delimited, record termination strings, &c.).My problem is that I have yet to find the equivalent within MS SQLserver. I did find a LOAD statement in T-SQL, but at first glance itseems to do something completely different.How does one normally load data from a plain text file into a table inMS SQL? This needs to be relatively efficient since, once inproduction, it will be used to load tens of megabytes of data into thedatabase (a feed from a data provider). Is it flexible enough to allowme to specify whether the fields are tab delimited vs comma delimited,optionally enclosed by quotes, record termination charactors, &c.?All I really need is direction to the right part of the T-SQL reference(MS SQL Server 2005). Anything else, such as examples, is icing on thecake.ThanksTed
I'm using the For Each loop container to load multiple XML data files into SQL Server, and noticing some peculiar behavior and need some advice.
The pattern I'm trying to accomplish is this: Iterate over a collection of XML files in a specific folder, loading each in turn into SQL Server. If the file has already been loaded, delete the records first before the load. After the load succeeds, move the file into an Archive folder.
To accomplish this, I've set up a For Eac Loop container using the For Each File enumerator, and retrieve just the file name and extension into a variable. The first task in this is an Execute SQL task that uses a SQL DML statement to delete records based on a field in the table containing the file name (DELETE FROM table WHERE PROG_NAME = ?), and map a variable to the parameter. The next task is the data flow task that uses an XML source using the variable as the file name, and SQL Server destination. I use a derived column task in between to plug the variable holding the file name into the PROG_NAME field. So far, so good. This works.
But now comes the peculiar part. I initially had the XSD files in the same folder as the XML files, but wanted to put them in their own directory, so moved them, and made the change to the XML source adapter for the new path to the XSD file. The next time I ran my package, it failed. For some reason, as the For Each Loop tried to iterate over the directory, it was using the XSD path assigned in the XML source instead of the path for the XML files. Unusual...
My question is, why when choosing the File name & Extension retrieval type (as opposed to the fully qualified name) will the task try to use the XSD location to find the files? Is my variable getting reassigned somewhere?
hi , I am trying to copy files dynamically from a remot folder by using File System task , the files are going to be generated daily with f1yyyy-mm-dd.txt I created 3 file connections : - Existing remote folder (RemotFolder) and set that to var v_remotefolder - Existing destination folder (DestFolder) - existing file (filename) and set the expression to var (string ) =v_fname in the File System task I set the the Destination connection = DestFolder Operation =Copy file Source = filename expression >>Property (source) = @[User::v_remotefolder] + @[User::v_fname] I know I need the date too ,but I 'm getting error on this step. Can you please tell me if you know how to resolve this without using For each loop since I don't have control on the remote folder and I don't want to copy unneeded files (the files are 10 GB daily) and get delete it once every month. -I know this not uncommon task but I can't find any example. Thanks
Here's my delema, I have a file that's 308 bytes wide by 5.7 million records. The record length is fixed and the position and width of the known within the record. When I run DTS I recieve this error Msg MS DTS flat file provide and Err Diesdription: error creating file mapping view: not enough storage is available to process this command. Then when I try to continue with the wizard, it will not allow me to separate the data into the format that I need. Is there any other way to import this file using DTS?
I have 6000+ text files, average size 400 kb, that I need to load into 1 table in Sql Server 2000. Does anyone know of an easy way to do this? I thought I would just write a little VB app to loop through all the files in the directory and insert the data into an existing table but there must be an easier way.
Currently we are trying to load the xml files into sqlserver tables by using ssis 2012,We are getting xml files as a column in source table ,so we have to push these xml files into destination tables.
I'm following the below way to perform this activity
[URL]
But We have standard XSD structure for all the xml files ,and if xml file matches the XSD structure then only we have to load ,else it should skip to next xml file.
I need to get a list of customer ids and then use them as a parameter to select from a transaction table and then create a file for each customer. I have used an execute SQL task to get the list of customers and have put the result set into a variable.
How do go through the recordset to create a file for each customer?
Hi.I need to give my customer an sql file that they can run in query analyzer.All the stuff they need to run is in a set of existing files.I'd like to just tell them to load this file (this is oracle syntax):@file1.sql@file2.sql@file3.sqlis there some way of calling these files (that are in the same dir) from amaster sql file?ThanksJeff Kish
I've just started using SQL Server 2005 Integration Services and I've come up against a situation where I need to name output files dynamically (i.e. based on a timestmap). Looking through this newsgroup, and other web resources it looks like my only option is to use a Script Task/ActiveX Script Task to rename the file after it has been created with a generic file name. I was wondering if there was a different approach or if there was a way to pass in a variable to the file connection manager that I could append to the file name at run-time.
How to download files from a webpage before loading into SQL Server tables? I have the following URL and under the Downloads & Resources section, I have different file formats.
By doing hover on the download tab for each file type, I see that there is a link that is associated with it just like the following:
For CSV - [URL] .... For XML - [URL] ....
The above is just an example for your reference/understanding. In the sample data from the internal website I have, I need to do a similar operation. The only difference would be that I would be having multiple XLS files with a description for each.
<li> <sub>Sales for Calendar Year 2015--All Countries </sub> <a href="/Data/Downloads/Documents/Sales/Sales_Quarter1.xlsx"> <sub>[XLS]</sub></a><sub> , <a href="/Data/Downloads/Documents/Sales/Sales_Quarter1.pdf"><sub>[PDF]</sub></a><sub></sub></sub> </li>
I need to download the file based on the month/quarter every time.
There is one manual process done by my friend. The process is " In query analyser, she runs a sql query and the resultset of Sql query, she copies and pastes into a excel workbook".
I wanted to write a SSIS package to automate it and providing a UI, in which she can give connection details and paste the SQL query and the package will in turn generate excel workbook in the target path". I thought of trying ExecuteSQL task, but the output resultset of ExecuteSQL task cannot be mapped dynamically to excel source(correct me if I am wrong). If I use Dataflow task, then also whether dynamic mapping of SQL output to Excel destination input is possible?
Can you please tell me the best approach for achieving the above functionality?
Hi there - can anyone advise on the following issue. We have recently performed some server side tracing on a particular SQL instance over 24hr period. We are now attempting to load these into a database for analysis. Here lies the problem.
When we are loading the profiler trace files (one at a time) into the database the transaction log is growing at an excessive rate. Even though the database is in SIMPLE mode.
We are loading the traces using the command:
INSERT INTO sqlTableToLoad SELECT * FROM ::fn_trace_gettable('MytraceFileName', DEFAULT)
Can anyone advise how we could possibly get round this issue as we're running out of space due to the transaction log.
Every day an application creates new tables and dumps static info into them.
I would like to create a package to dynamically export those database tables to raw files for long term archive, one file per table. Here is what I have so far and the issue I am having.
1) Get a list of un-archived tables. 2) Foreach table do the following.
a. Export the table into raw file. b. Zip the raw file. c. Update archive tracking table.
As long as the metadata for each table is the same this package seems to work fine. However, I have many tables with different metadata. How can I dynamically get the package update the metadata column collection when it hits a new table? When it hits a table with different metadata I am getting warnings like this:
The column "some_column" needs to be added to the external metadata column collection.
The "external metadata column "someother_column" (103)" needs to be removed from the external metadata column collection.
Then I get this error: Error: 0xC004706B at dump the table into a raw file, DTS.Pipeline: "component "OLE DB Source" (1)" failed validation and returned validation status "VS_NEEDSNEWMETADATA"
I have developed an SSIS package which extracts and creates 5 flat files and finally using Process Extraction task zip the folder. On my Dev environment everything is working fine but when I am moving to SIT and UAT, not able to set up jobs dynamically by importing XMLConfig file.I created variables and assigned values but still it doesnt take.Below are varaibles I created for flat file destination, Arguments and Working Directory (for zipping)On UAT when I go to SQLAgentJobs to set, import .dtsx file, XML config file....the new values doesnt appear. why ?DataSource is taking always dev location....why ? How can I set it up to take dynamic values what I mentioned in config file ?
I have a load (180,000+) of text files whose contents need to go into a SQL server database.Whats the best way of doing this? Using a c# console program and if so, using FileStream or StreamReader? Or using a feature of SQL server itself. The text files themselves are less that 1k and are literally less than 200 characters.The problem is, I've tried a WinForm and although I can detect what files are there, as soon as I attempt to open one for reading, everything stops working and won't insert anything to the database.
Hi allCould someone help me with the following problem? Hours of googlingyesterday couldn't get me the answer. I'm using SQL 2000 and DTS andtrying to import a huge fixed width text file.File is >1m rows and >200 columns and is defined by a proprietory (i.e. notbcp produced) format specification of the formName Start LengthFld1 0 20Fld2 19 5Fld3 24 53and so on.Tbe only way I've found to define the columns so that DTS can import thefile properly is to go through the wizard and click on the starts of eachcolumn. I don't want to use bcp if possible (I did enough of that onSQL7) - but surely there's a way to get DTS to read from a format file so Idon't have to click 200 times (with all the ensuing errors I could make).Any help greatly appreciated.CheersRob
Hi!I have a large project that is due to complete this week. In order tocomplete it I need SQL Server 2000 installed on a remote server. Mydisk is corrupt and to order another media disk would damage mydeadline. I have the licence and serial key, but now need good installfiles. I am even ready to buy another retail box, if I can find asupplier that would give me a download site for the media, while I waitfor the shipment!Please PLEASE help!Regards,Barry
I built an application that connect to MS SqlServer2005 using Native driver (sqlncli.msi) I install that file from MS site, I need to deply my application to the end-user, and I would like to know what files do I need to deploy to make sure the application is gona run okay on the client PC's.
I search in the registry for the driver, and I found this "sqlncli.dll", is it enough or I need to include more files !!
I've been writing this stuff for a while, and can't seem to come to the conclusion of how I should be retrieving data and assigning this data to variables.
Since i'm using SQL Server, I'm convinced that I should be using the datareaders GetSqlDouble (or whatever) function, but this would mean i need my local variables to be one of the SQL types. The problem with that is, that there will have to be lots of conversions done by me to be able to use a SQL type in my application.
For instance, I have a class where i'm retrieving dates. In order to retrieve them correctly (Null values included), I need to retrieve them with GetSqlDateTime(), then when it comes time to display the date in a table, i must first check for nulls, then convert to a string. This seems to be very cumbersome. Would I be better off just using GetDateTime(), and the .ToString method, and ignoring Sql Types all together?
so, basically, how are you guys using your sql server data? with the supplied sql types, and doing all of the post-processing work manually? I feel like i'm having trouble conveying my issue...hopefully someone knows what i mean....i'd just like some direction to save trouble in the long run, since i feel like there's got to be a better way...
