SAS Code abend with RC 08- 'File is full and may be damaged&

Subbulakshmi yogesh · New User Joined: 28 Sep 2011 Posts: 8 Location: india

I am try to run a SAS code to read an input file and do a process filteration and create an output.
my input file is with 50 LRECL ~16 million records volume. My output file is 40 LRECL and expect some 1 million records out. The SAS work i mention around 8000.

//STEP01 EXEC SASV9,REGION=0M
//SAS.WORK DD SPACE=(CYL,(8000,8000),RLSE)

But i am getting the below error.
ERROR: Write to SASDATASET.DATA failed. File is full and may be damaged.

i just tried increasing the SAS workspace and also gave the REGION parameter. But still my job is failing. Can anyone suggest me to solve this abending SAS step. we can create a permanent dataset for the intermediate SAS step, but i m bit worried about the DASD allocation. Is there any other option to rectify this.

expat · Posted: Fri Oct 11, 2013 1:05 pm

What exactly is this filtration process ?

You may find that your SORT product may do exactly the same thing as your SAS code.

Unless you give us some usefull information all we can do is sit here and ignore your problem.

Unless your WORK dataset is allocated a special STORCLAS and/or DATACLAS such a large space allocation may well be ignored and set to s pecific default. This is entirely site dependant so we can not help with that aspect of the problem. I would try (CYL,(1000,1000)),VOL=(,,,5)

Please show the code that you are using and the error messages issued in full. Also any IEC error messages that may have been displayed.

SO - you give us the information requested and we may be able to help you.

And on a purely technical point, the job DID NOT ABEND RC 08, it merely issued a return code 8 from the SAS step.

PeterHolland · Posted: Fri Oct 11, 2013 1:53 pm

As far as i can see, it is not your WORK dataset but your SASDATASET dataset (your final destination) that is too small. So define a new target dataset.

Subbulakshmi yogesh · New User Joined: 28 Sep 2011 Posts: 8 Location: india

i got my issue resolved. i created the SAS dataset as,

DATA _Null_;

as i dont need to set this data anywhere. this job completed fine. thanks for all your points.

expat · Posted: Fri Oct 11, 2013 3:05 pm

Peter,

You may well be right, but usually the error message is DDNAME.SASDATASET.DATA and I for one also thought that it may have been the destination dataset.

But as nearly always only incomplete information is given leaving us to guess and offer possibly unsuitable solutions, wasting our time.

So until we get some real information from the OP we can't really do too much else apart from the usual .....

PeterHolland · Posted: Fri Oct 11, 2013 3:25 pm

Hello expat,

if a libname is used the allocation is dynamic. SAS only displays the libref name in this case SASDATASET and the name of the library member.

For the rest of your comment im with you.

Subbulakshmi yogesh · New User Joined: 28 Sep 2011 Posts: 8 Location: india

SORRY. maybe i didnt illustrate the problem clearly. Actually my work dataset is set with some specific default allocation. As my input file exceed that limit i m getting the FILE is full message.

i just tried allocating a permanent dataset and assign that to the SAS dataset it work. but actually i dont want to do that way.

i just solved my problem with _NULL_ which process the datasets with the actual DATASET creation.

expat,

my filtration includes a process of comparing the date in the input file with the SAS TRANSPOSE option and it cant be achieved with SORT.