whether row hash changes when new nodes or Amp added?

Database
Enthusiast

whether row hash changes when new nodes or Amp added?

Hi All,
As I am going to prepare for PI and puzzled in few questions, looking to find the answer of these questions...

whether hash row changes when we add new AMPs or Nodes in the existing Teradata System....?

can any body put more lights on it....

Thanks in Advance....

6 REPLIES
Enthusiast

Re: whether row hash changes when new nodes or Amp added?

If you do a reconfig, you'll get a new Hashmap
Enthusiast

Re: whether row hash changes when new nodes or Amp added?

The hash map is recomputed when the AMP configuration changes, but the hashing function itself is independent of the configuratiion. In other words, row hash values do not change.
Enthusiast

Re: whether row hash changes when new nodes or Amp added?

RowHash is getting generated from the Primary Index value. Given the Primary Index value to the Hashing Algo, it will generate the row hash. So, RowHash is always dependant on the primary index value, and it has nothing to do with the number of AMPs. However, whenever the number of AMPs get increased/decreased, the HashMap size is getting adjusted accordingly to accomadate the currect number of AMPs.

Hope solved your doubt.
Enthusiast

Re: whether row hash changes when new nodes or Amp added?

hi
thanks all for replying.... it is more clearer to me....

Thanks......

Re: whether row hash changes when new nodes or Amp added?

hi

Does Row hash for a PI value changes as and when its datatype changes?

Supporter

Re: whether row hash changes when new nodes or Amp added?

Yes for real data type changes like char to integer.

And smallint and integer values are hashed the same (for smallint range)

Check the outcome of hashrow 


select hashrow(cast('123' as char(10)))


select hashrow(cast('123' as varchar(100)))


select hashrow(cast('123' as integer))


subt["select hashrow(cast('123' as smallint))"]