DSXchange

hello, you know my habit of not opening new posts since similar problems are already reported in here. right now I am facing exactly the same problem described above. I have a dataset with a column MyFlagCol which is all set to value 'N'. the data is then entering an aggregator where I use MyFlagCol as a group key. but after the aggregator stage MyFlagCol randomly has values of "N" and "" (empty string or null, I couldn't figure it out very well from the peek). obvious enough I am expecting to get an "N" for all values but tough luck cause it aint happening.

to answer ray's question my data is not sorted and uses partition Same.

if you have any advice I would gladly take it

Same doesn't help. What's it the same AS ? That is, what is the upstream partitioning algorithm that's being used?

good point Ray the dataset is actually the first stage in the job so Same would not be very helpful ... hmmm I dunno, I would chose Entire then but I am afraid not to mess up the data ..

teddycarebears wrote:hello, you know my habit of not opening new posts since similar problems are already reported in here.
<snip>
if you have any advice I would gladly take it

First bit of advice? Break this habit. Similar != same, so please start a new post and then reference the old post if you feel it would be helpful in solving your problem. That way we get the added bonus of knowing your version and type and O/S plus you get the bonus of being able to flag it as Resolved if everything works out.

Split from this older and similar topic, btw.

hello there, thanks chullet, good info to know for the future

yes, Ray's answer helped in a way because it made me look at partitionings, field types and sortings all around the sequence and with a few minor modifications of sortings and nullabilities everything works now like a charm.

so thank you all