Error: Failed to keylookup for record

stiwari · Post by **stiwari** » Wed Aug 30, 2006 4:46 am

Hi All,

I am trying to load 120 million record to my fact table....instead of using join i am using lookup to join all the dimention tables....but after loading 40 million record its getting aborted.

Error message is :- Failed to keylookup for record

I am not able to understand what error it is....can any one plz suggest me what could be the reason?

Thanks in advance
stiwari

ArndW · Post by **ArndW** » Wed Aug 30, 2006 6:13 am

You are using a key for your lookup and the error states that it didn't find a lookup record for that key and you have set the job to abort when this happens.

kumar_s · Post by **kumar_s** » Wed Aug 30, 2006 6:25 am

You can avoid aborting the job with this message by removing the option of abort if no match. But ultimately you need to find the reason why you didnt have the match for fact over the dimension table.
Hope the replies for all of your 4 post might have given you an idea that by doing a search on this forum, you can find the solution for your issue.

stiwari · Post by **stiwari** » Wed Aug 30, 2006 8:21 am

Thanks guys....ya, I got solution of all the issues. I got very quick replies with the valuable solution.....i mean the people on this forum helped me a lot.
thanks once again to all the Datastage guru's :D

stiwari

stiwari · Post by **stiwari** » Thu Aug 31, 2006 6:14 am

Hey Guru's

I just wanted to know the functionality of the constraints of the lookup failure......for the DROP......is it drops those records, those doesn't match with the joining condition or somthing else......?

stiwari

ray.wurlod · Post by **ray.wurlod** » Thu Aug 31, 2006 6:05 pm

Is your F1 key broken?

stiwari · Post by **stiwari** » Fri Sep 01, 2006 2:36 pm

That's what I pressed to get help from Datastage Guru's....

, Actually I searched it but didn't get the proper answer....can you plz explain me...about this property?

stiwari

ray.wurlod · Post by **ray.wurlod** » Fri Sep 01, 2006 9:21 pm

"Drop" causes any input row for which the lookup fails not to appear on the output of the Lookup stage.

By discarding records in this fashion you are eliminating any chance of performing row count reconciliation. Prefer a rejects link from the Lookup stage, even if it appends to /dev/null. At least this way all input rows can be accounted for.