DSXchange

The reason is happening is due to the fact that Sybase Open Client connectivity used 64-bit architecture while the ODBC driver is designed for 32-bit architecture that is used by Datastage.
Hence this instability & non consistent behaviour.

Regards,
NV

If you trying to do Insert/Update on a Sybase table with default Page Level locking, then this is possible.
In that case u need to change the table lock to Row level.
:lol:
NV

Yes, even I had implemented this in a Server job. Here I am using a single transformer stage having 2 output links, one for delete & other for Insert and executed in that order. But No Luck. Anyways,I have now split my job in 2 parts..one will Delete and then another will Insert using the DS cre...

Although doing this may slow down the job, after doing that too, it doesn't work.It still hangs up.
Any thing else I can check for?

HI,

This table can be populated by other processes too on the same day...we want to delete only those rows that were inserted by our process.. SO using only the dates for delete without the keys will not help.

Has it got something to do with table locking in Oracle..Row Level vs Table level?

NV

Hi, We have a requirement where we need to clean the table prior to insert. Example if src has 100 records. The idea is, if for some reason the job fails/aborts/hangs (may be due to DB problem or any other reason), when we RE RUN the job it should clear the previous loaded records (if any) and then ...

NO..Fix_Zero is for decimaltostring conversion..

I have tried doing this conversion in a stage varibale and then used the SV in the derivation..

Hi, I am reading "amount" from a fixed width seq file as a Char e.g (2259 or 36153.35) and then the only transformation I do before writting to an Oracle table is do type conversion using StringToDecimal. For some strange reason it stores these values as 2259.580000000000 and 36153.3499999...

Hi,

I had already tried it, but it still throws the warnings. It seems writting a Unix script for this might be the only option.

Thanks,
Nirav

Thanks to All..
I will test it out

Thanks chulett.
This is precisely what I tried and it worked.

Thanks to all for ur reply. 1. We do not need "dot" as part of the file name..that would come anyways as with all file names..we need only B or C dynamically. 2. Yes, I am already doing it using a parameter..but I was thinking if there was any other way where we could directly do it w/o us...

Hi, we are using a multi instance job and will be passing values like B and C as invocation ids so that we can read from different source files. I want these to be concatenated to the final outbound files respectively.So what should I add as a part of output filename to be able to do it. e.g. we wan...

Hi,

We do not get any record type identifier.We simply get a header rec as the first record and the Trailer rec as the second last record with an empty line as the last one.

Hi, I get the following warning message when I try to read from the Source Sequential file "SEQ_Histpion,0: Delimiter for field "Field004" not found; input: {0d}, at offset: 107" The reason is that the Header (and also Trailer) record do not have same metadata as all other data r...