DSXchange

Hi,
I have developed the job in which i m using Lookup stage,but the job is getting aborted and giving error message is:-

Lookup_1,0: Could not map table file "/etlstg/datasets/lookuptable.20090506.hlpp0hd (size 541052504 bytes)": Not enough space
Error finalizing / saving table /tmp/dynLUT87267436f27d22

My config file is

main_program: APT configuration file: /dshome/Ascential/DataStage/Configurations/etlstg_dataset.apt
{
node "node1"
{
fastname "CNDAIRPZDBDT03"
pools ""
resource disk "/etlstg/datasets" {pools ""}
resource scratchdisk "/dshome/Ascential/DataStage/Scratch" {pools ""}
}
.................................
At resource disk mountpoint there are around 220 GB of space is free but still my job is getting aborted.
Please provide me any possible solution.

Thanks in advance

How many rows and how large is each on the reference input?

I think this is related to the memory space and not really the disk space. Try using a join stage instead of lookup....

ray.wurlod wrote:How many rows and how large is each on the reference input? ...

there are around 6kakhs record in the reference ........and i can not use join stage becoz i want the rejection.. and i dont believe on merge stage....i have seen that merge stage giving worng result...so plz provide me solution for this..??

You may use a Left Outer Join. So the unmactched values may be rejected in the following filter stage.

prasson_ibm wrote:and i dont believe on merge stage....i have seen that merge stage giving worng result...so plz provide me solution for this..??

Define "wrong result"

wrong result means data in the target dataset inproper i.e. not matching with the source dataset

If you were not getting the desired result that doesn't always mean that the stage is not working correctly. It may be due to wrong logic.

If you are sure it dosen't work the report it as a bug with IBM with sample.
Well. it works fine for me.

You can avoid putting unnecessry columns to the reference link. If you have already done that then you can always use join stage with left outer join and downstream filter to get the rejected records.