waitForWriteSignal(): Premature EOF Error

Post questions here relative to DataStage Enterprise/PX Edition for such areas as Parallel job design, Parallel datasets, BuildOps, Wrappers, etc.

Moderators: chulett, rschirm, roy

Post Reply
g_rkrish
Participant
Posts: 264
Joined: Wed Feb 08, 2006 12:06 am

waitForWriteSignal(): Premature EOF Error

Post by g_rkrish »

Hi,

We are upgrading from 7.1 to 8.1 we have encountred a problem on the oracle DB where there is user defined query in that which does some join etc going to a copy stage but it throws up an error saying

waitForWriteSignal(): Premature EOF on node xxxxxxxxx Socket operation on non-socket

but the query runs fine from the database but i try to view the data from the Datastage jobs it throws out an error saying

When binding output interface field "field1" to field "field": Converting a nullable source to a non-nullable result;
a fatal runtime error could occur; use the modify operator to
specify a value to which the null should be converted.


but the metdata in the oracle database is same as metedata in Datastage though.

any thoughts on this
RK
ray.wurlod
Participant
Posts: 54607
Joined: Wed Oct 23, 2002 10:52 pm
Location: Sydney, Australia
Contact:

Post by ray.wurlod »

They are separate issues.

The first - is it intermittent or permanent? This one can occur if connection is lost even for a short time.

The second has a fairly self-explanatory message. Check both lots of metadata again, including re-importing the table definition (maybe to a different folder) - "they" may have changed it.
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
g_rkrish
Participant
Posts: 264
Joined: Wed Feb 08, 2006 12:06 am

Post by g_rkrish »

The first one happens whenever we run the job but the in the second one there is no nulls in the specified field but it throws up an error.
RK
ArndW
Participant
Posts: 16318
Joined: Tue Nov 16, 2004 9:08 am
Location: Germany
Contact:

Post by ArndW »

I would fix the nullability issue first. Your "field1" is declared as nullable in your database - check your table DDL to confirm. Once you handle the null then you can attack the second problem, if it still exists.
g_rkrish
Participant
Posts: 264
Joined: Wed Feb 08, 2006 12:06 am

Post by g_rkrish »

ArndW,
Database metadata even with nullablity is identical with the Datastage. Is some thing to do with datastage versions.
RK
ArndW
Participant
Posts: 16318
Joined: Tue Nov 16, 2004 9:08 am
Location: Germany
Contact:

Post by ArndW »

You should check your database DDL, not (only) the DataStage definition.
g_rkrish
Participant
Posts: 264
Joined: Wed Feb 08, 2006 12:06 am

Post by g_rkrish »

I checked it is same.
RK
ArndW
Participant
Posts: 16318
Joined: Tue Nov 16, 2004 9:08 am
Location: Germany
Contact:

Post by ArndW »

If your DDL and database read stage both specify a non-nullable "Field1" then I would add "$OSH_PRINT_SCHEMA" to your job parameters and see what the actual schemas that DataStage uses are. If they differ from the database DDL then you should submit this issue to your support provider.
g_rkrish
Participant
Posts: 264
Joined: Wed Feb 08, 2006 12:06 am

Post by g_rkrish »

they are coming correctly but i feel i use a case statement in the select with that field, so will Datastage takes it as nullable by default even though the source has nonnullable and have a default value as Decimal(38,10) even though the source has integer which is int32.
RK
ArndW
Participant
Posts: 16318
Joined: Tue Nov 16, 2004 9:08 am
Location: Germany
Contact:

Post by ArndW »

Why not declare the field as nullable, then do a dummy NulltoValue() or similar call downstream?
g_rkrish
Participant
Posts: 264
Joined: Wed Feb 08, 2006 12:06 am

Post by g_rkrish »

That i checked earlier with making it as nullable it was working .but why is the problem comes when i use NVL in the query to substitute, is with Datastage 8.x assuming the computed value as nullable irrespective of definition.My concern is we are doing upgrade with the DataStage as we have tons of jobs it affects every job where we used which eats up time can we do any fix for that other than changing code. NVL works fine with DS 7.1.
RK
Post Reply