DSXchange

Can we have a job with 4 different database stages (API or DB2 connector) from transformer and we will be updating same table in all the 4 stages. As this may cause table locks, I thought If I order the links one by one that would take care of the lock issue but one incoming record may send data to ...

I found out a way to not create many files by using transformer and updating directly on the table but this may cause table locks... Can Link Order resolve this issue ? In this situation one record may go to more than one update links. How does the link order works ? Does it hold the record for 2nd ...

I am configuring SCD stage but would like to update all prev versions for Type-1 columns and have few of them in the table. My plan is to map columns from both source and reference for the Type1 columns in Fast Path 5 of 5 and then write to a Fact File not the Dim File. In next job, compare both bef...

Yes. You are correct but in my case it's a batch process and we won't get two records for same business key in a load as we extract from OLTP systems at certain time in a day. Does it make sense ?

Ray - Is not what you are describing is true If I Insert directly into the Dimension table in the same job from SCD stage ? What difference does it make if I just load them to the file and do bulk insert and update as I have to update all the history record with Type1 values not just the last one. P...

Sura - I got it...Seems like I have interchanged the links... Another question is now we have to UPSERT into the Dim table. Correct ? But one of the best practice says that DO NOT DO upserts as it has to check for every record to make sure if it's Insert or Update. So how does it affect the performa...

I am configuring SCD stage in 8.7. Job does not fail but it does not give me the expected result. For testing purpose, I have changed a Type-2 column and ran the job which only generated 1 record in the DimUpdate link which is Insert record. It did not create the record for corresponding update(to c...

Ray,
Thanks for the reply. We have moved from 7.5 to 8.7 so need to play with SCD stage. Never used it before...I will try that...

I am wondering what is the best way to deal with Hybrid SCD types in a single table in Datastage 8.7 ? When I say hybrid SCD types, I meant, I don't need to create a new version of a record for any column in a table changes rather only if certain column changes otherwise I just need to update the re...

We are having this weird issue where a simple job fails with following error. Abnormal termination of stage Job_Name detected. Reading from a seq file and looking up to a hashed file and writing to a seq file. Problem is that if we have "Preload to memory enabled" then the job fails but if...

I figured it out. It's because of the null value in the transformer...

Thanks

Seq File --- Transformer -- DB2Stage API stage. I get the following error from Sequential file stage. sqf_States,0: Fatal Error: Unable to allocate communication resources It's a new project and this is the first time we are trying to load a table in this database. I tried few options mentioned in t...

I think as someone mentioned above it does not matter where it exists on the pallete. If it changes the number of records/columns when it pass through the stage then it's active otherwise passive.

Ray - Yes. Join stage does support more than 2 inputs but we can't do full outer join between all the inputs.

I am wondering if there is another way of doing full outer between 3 or more files. Now, I use join stage to join two files and the output will be joined to the 3rd file. Is there a better way ?