| cdaringer |
| Newbie |
|
| Earth |
|
|
| None Specified |
|
| Sunday, March 10, 2013 |
| Sunday, March 17, 2013 4:46:44 PM |
2
[0.01% of all post / 0.00 posts per day] |
|
I will of course send an email to support on clarification on this topic However not wasting time trying to follow the documentation is sound advice:
Using Oracle ODBC everything works fine via the altova interfaces:
But following the instructions carefully in the documentation can get the sun.jdbc.odbc.JdbcOdbcDriver connections to work as a Bridged Connection with a System DSN and then the connect to work fine, however the Retrieve Data shows an "Execution in Progress" and nothing else, not even a red-button to cancel the job on the server.
I have been one of three people to look at this and following your documentation carefully and cannot get it to work
However:
I go to Eclipse and use jdbc:oracle:thin:@localhost:1521:orcl no problems however I'm referencing ojdbc14.jar from the dbhome of Oracle
but things work like a champ..
http://manual.altova.com/mapforce/mapforce-professional/index.html?mffjdbc_driver_setup.htm
No mention of ojdbc14.jar in your documentation at Altova
What is not happening is seeing any other driver selection than the sun.jdbc.odbc.JdbcOdbcDriver
What of course would be beneficial is if the documentation had the corresponding configuration requirements which do reference:
jdbc:oracle:thin:[user/password]@//[host][:port]/SID
jdbc:oracle:thin:@//abcd234/ORA11
F.Y.I. all of my tests were done in VM on both XP and Win7 with a Database local to the VM running in one test and in a separate test where the Dev Days Oracle database was run under VirtualBox with Redhat and the Altova on Win7 and XP on VMworkstation 9 within eclipse: "JDBC:10.2.0" is being referenced.
Why this is of course a bit confusing/difficult is that look-up tables in mapforce would of course use a common initialization at development and when creating Java code for use.
However I agree Vlad that it is a waste of time addressing this issue here
|
The product is pretty dynamic but comes up short in last mile functionality in two areas:
First is the ability to map using "Filter node-rows" conditionally to two fields in the textfile output of fixed length and then treat the empty fields as absent.
If my mapping logic said on Monday send free coffee coupon to Vlad in length of 10, and Tuesday send free coffee coupon to Paul in length of 15, and I mapped the coupon data to two fields in the text-file and ran-conditionally on the day of the week and selected fixed length output and treat empty fields as absent.. the requirement to meet both lengths is not met. Somebody doesn't get the free coffee coupon without allot of additional effort
Second is the ability to map using "Filter node-rows" conditionally to a single EDI object and provide conditional delimiters so you could have your coffee delivered! In one scenario Vlad prefers vendor A to deliver his coffee and that vendor uses their coffee edi with one implementation of:
ComponentSeparator
DataElementSeparator
SegmentTerminator
ReleaseCharacter
Paul prefers another vendor! His vendor uses a different
ComponentSeparator
DataElementSeparator
SegmentTerminator
ReleaseCharacter
The lack of an ability to using "Filter node-rows" conditionally at run-time for setting serviceChars makes getting your coffee delivered more difficult
These scenarios are the only two short-comings in the product in my opinion..
It would be neat if you guys started thinking about data pairs for hadoop.. and see some objects for map-reduce and hadoop as the GUI is good
But those two last-mile items make simple scenarios difficult
|
|
