Error while binary DUMP.

[Arunselvan]

Hi all,

I am dumping my database using proutil command. Every table gets dumped without any problem. But one particular table shows the following error.

02/01/14 21:38:01: Start Dumping sod_det of /qad/qaddb/dbname to /qad/qaddb/dump/
OpenEdge Release 10.2B07 as of Fri Sep 7 02:16:34 EDT 2012

Using index 1611 for dump of table sod_det. (6101)
Could not find record. Recid 115150942, error number 8, return -1. (12011)
Binary Dump failed. (6253)
02/01/14 21:38:08: Finish Dumping sod_det of /qad/qaddb/dbname to /qad/qaddb/dump/

Please suggest some solutions. It's very urgent.

Regards,
Arun.

 

[oli]

You are apparently using a corrupted index to dump your table.
Use another index (by providing the index number to the proutil -C dump) , or try first to fix it using idxfix or idxbuild.

 

[Arunselvan]

Hi oli,

Thanks. It works. And I have one doubt.

while dumping will it use one index definitely? if it would use, then please tell me why? and what will happen when we dump the table with no index?

 

[Rob Fitzpatrick]

By default, binary dump uses the primary index. As oli indicated, you can direct it to use a different index; it cannot be a word index. Also, if the primary index is inactive a different index will be used. If no active indexes are available the dump will fail. The dump index determines the sort order of the records in the dump file.

If you specify a no-index dump by using the "-index 0" parameter, the behaviour is determined by the characteristics of the storage area where the table resides. If it is in a Type II storage area (blocks/cluster > 1) then the dump will be performed without an index and you will see a message like this in the output:

Performing table scan for dump of table customer. (14653)

In this case the records in the dump file are essentially in random order (their order on disk).

If the table resides in a Type I storage area (blocks/cluster is 1 or is not specified) then the "-index 0" parameter will be ignored.