Kbase P89971: Bulkload fails to load data file with error: unable to read the .d trailer (1818)
| Autor |
Progress Software Corporation - Progress |
| Acesso |
Público |
| Publicação |
01/03/2010 |
|
Status: Verified
SYMPTOM(s):
BULKLOAD fails with error could not read the .d trailer
Warning: Unable to read the .d trailer. Using <code-page> as the code page for conversion. (1818)
Warning: Unable to read the .d trailer. Using ibm850 as the code page for conversion. (1818)
Progress 8.1A03
Data Dictionary dumped a 3.9GB .d file.
FACT(s) (Environment):
Progress 8.1A
CAUSE:
Reached the Data Dictionary limit of 2GB on reading a file during the bulk load process
FIX:
Load the data file using the proutil bulkload utility by creating files less than 2GB:
1. Split the .d file that is over 2GB into multiple files
2. Limit each file to less than 2GBs
3. Use a unique name for each file
4. Do not use the current trailer as constituted
6. Copy the trailer and delete all references to record counts and the character count, i.e:
.
PSC
filename=Customer
ldbname=sys
timestamp=2004/08/02-15:43:31
numformat=44,46
dateformat=mdy-1950
map=NO-MAP
cpstream=ISO8859-1
.
7. The dots "." before and after the trailer need to be included
8. Copy the section of the bulkloader description file(.fd) that references this table, i.e:
Customer customer.d customer.e
CustNum
Country
Name
Address
Address2
City
State
PostalCode
Contact
Phone
SalesRep
CreditLimit
Balance
Terms
Discount
Comments
Fax
EmailAddress
.
9. Create a number of .fd files to match with the number of .d files
produced on the split, each a copy of the original section for this
table, i.e. dbname2.fd, dbname3.fd
10. In the original.fd file change the name for this .d file to the unique name
given to the first .d split off for the table.
11. Run proutil dbname -C bulkload dbname.fd
12. In the remaining .fd files change the .d file to the name for corresponding
data file.
13. Run proutil dbname -C bulkload dbname2.fd
14. In dbname3.fd rename the .d file to the name for the third file of the split
15 Run proutil dbname -C bulkload dbname3.fd
16. Repeat as necessary base on the number of .d files created in the split.
17. Rebuild the indexes.