I am trying to access a text file and I am getting the following error message:

An error has been detected during server to client data transfer.
ERROR: -12: Pcb2.: java.io.EOFException

The text file has over 20,000 records. When I use a smaller version of the file (approximately 100 records) it works perfectly. My understanding was that WebFocus could handle millions of records, and I have opened files as large as this one with a different master file.

Any help would be greatly appreciated.

Are you possibly timing out?
We set our time out to 10 minutes (default was 5 minutes with install)

Just a thought!

It doesn't seem to be timing out. The message comes back very quickly--within seconds.

Could it be a data value in the larger text file?

Have you tried a RECORDLIMIT, starting with a low number and then increasing it until the problem occurs?

Welcome to the club !!!
I had a similar problem for the last 2 months and just last week got it resolved. Please visit my posts on this topic for more details on the nature of the problem

How it was resolved???
**********************
this one was suggested by WebFOCUS support
**********************
It involves setting the agent_refresh parameter on the
server to 1 in your EDASERVE.CFG file. This causes the agent to refresh and
the edatemp directory to be cleared after each connection. It is possible
that this directory is currently filling up until the current agent connection
number is being hit the default is 100, and if this number is currently set
to a high value, it may not be hitting this limit but just filling up the
directory.

To do this, edit the EDASERVE.CFG file on your server and set the value to
the following
or add the line if it is not currently there:

agent_refresh = 1 <-----Note: this entry is in lower case.

Then, refresh your server to pick up this change.

***************************************
check your code if you have something like this
***************************************
-IF &REF EQ �CT� THEN GOTO CTO;
-ELSE IF &REF EQ �WDTP� THEN GOTO WDTPO;

-CTO

Some code

-GOTO ENDCODE;

-WDTPO

SOME CODE

-GOTO ENDCODE;


-ENDCODE

When I changed the code to this the EOF Exception disappeared I did not know at this point that there were changes being done to the reporting server configuration

-IF &REF NE �CT� THEN GOTO CTO;

Some code

-CTO

- IF &REF NE �WDTP� THEN GOTO WDTPO;

SOME CODE

-WDTPO

******************************************
the third place where this could be replicated
******************************************
if you are using the same label name in 2 places
eg:

-IF &REF NE �CT� THEN GOTO CTO;
Some code
-CTO
- IF &REF NE �WDTP� THEN GOTO WDTPO;
SOME CODE
-WDTPO


some more code


-IF &REF NE �CT� THEN GOTO CTO;
Some other code
-CTO
- IF &REF NE �WDTP� THEN GOTO WDTPO;
SOME other CODE
-WDTPO

These were the changes we did.

Hope this helps

kalyan.

Lee,
I just started having the same problem against a file with Text. A couple things that may help....
set the LONGCHAR setting to Alpha and
if you create a hold file with the TEXT field make sure the TEXT field is in the last position.

TABLE FILE FILENAME
PRINT FIELD1 FIELD2 TEXTFIELD
ON TABLE HOLD
END

I'm crashing agents right and left. Kaylan, thanks for your post on adding the agent_refresh to edaserve, this may help?

Carol

AHA!
This line of code was crashing agents in WF 5.2.4

SET ONLINE_FMT = STYLED
This is an older command left in a rewrite of code.

Thanks.

I meant to say

SET ONLINE-FMT = STYLED

it's been a long week!

Thanks to everyone who has responded. I am having my IS department check on the EDASERVE.CFG file; I don't have access to make any changes to the reporting server.

Some things I have noticed: When I set the retrieval limits to limit the number of records it first would be run with 1212 records, but when I went back in and kept playing with record limit it pulled up to 2100 records. However, it doesn't really pull that many records. The file returned is only about 1000 records even though the source file has approximately 20,000 records. I am not applying any sort of selection criteria since I'm just trying to get it to actually read the records. The master file is a direct copy of an existing master file that my firm has successfully for a couple of years, and the date file was saved from a file created by that earlier master file, so the fields and data types should all be correct.

Lee,
We periodically experience this issue as well. There never seems to be a hard and fast reason. I reboot the server and everything that was broken then works.... I think it has something to do with the agents as well. I've tried setting mine up with other settings and it does seem we have fewer occurences of the problem but it hasn't completely gone away.

Lee,

I was also getting this error and was also advised by tech support to change the agent session refresh to 1. It worked!

I would really emphasize the need for this change to your WebFOCUS admin or whomever has access to the reporting server.

Does anyone know the difference between:

setting agent session refresh to 1 and
and
crashed_cleanup = on

Thanks!
Carol

agent_refresh = positive integer specifying limit for number of new connections


Internal default: 100


This keyword is part of the general workspace features provided in the scope of an agent service. It defines the maximum number of new connections which can be accepted during the life of each agent process. Beyond this limit additional new connections will be assigned to a fresh agent. Agent processes which have reached the limit will be terminated when the last accepted session disconnects.







crashed_cleanup = on or off


Internal default: off


By default, crashed agents are kept in the agents table for diagnostics purposes, and are only removed when they are explicitly killed. When this setting is on, crashed agents are automatically removed from agents list when the Workspace Manager notices the crash.

BTW. This info is available from the WebFOCUS Reporting server Web Console. I fyou go to Workspace, Configure, service and then click on one of the "?" it will bring up the page with details about the server settings.

kp,
Thank you for this information. I'm thinking we might need both settings. It seems if we have too many crashed agents, we see a random EDA/No Data.
I will check the ?

Thanks again,
Carol