on 11-30-2010 8:30 AM
Hi All,
I created the following Async scenario
DB2 -> JDBC -> PI -> RFC -> ECC
Fetching more then 150,000 records from the DB2 crashes the PI (the Java)
I was looking in the IBM site for solutions, didn't find any related parameters for the driver to configure.
How can I get the data without crashing the server?
Is it a known issue with the JDBC driver?
I need all the data - about 500,000 records every load
Thanks
Ranny
I already tried restricting the number of records, it works up to 75000 records then crashes.
Just for general knowledge the statement equal to "top(X)" in DB2 is
SELECT * FROM <TABLE> FETCH FIRST <XXX> ROWS ONLY
Thanks
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
To make sure I have a problem with the JDBC, I exported it to a flat file and changed the scenario to take it from that file.
As you can guess, there were no problems at all.
I tried MSAccess with ODBC, again no problems taking even more records from the DB2
So somthing with the JDBC adapter is making the server work till it crashs so hard I can't even start it again without stopping the JDBC service.
Since I'm taking the data from an external system, every change cost money, so I have to try everything before changing anything in the source system
Thanks,
Ranny
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Check if you can restrict the numbers of records by a DB specific select command.
I do not know DB2. In MS SQL it works with select top(5000) * from ...
Maybe you can also use a stored procedure to achieve the same result.
I assume that creating the XML structure from DB select result is too memory consuming, but I do not know how to prevent this behavior.
Hi Ranny
DB2 -> JDBC -> PI -> RFC -> ECC
SO u r using Sender JDBC channel.U need to configure it, based upon certain conditions...
your problem is no where associated to driver issue...
1. whether u r using update option, to filter out the previously read records.
2. Better use Stored procedure, rather then plain Select query,
Regards
Ashu
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
IBM site will not help you, you have to come with good design in this case, 1.5lac reacords means high volume,so insted of that pick up less number of records..like spliting..
Regards,
Raj
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
87 | |
10 | |
10 | |
9 | |
7 | |
6 | |
6 | |
5 | |
4 | |
3 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.