Hi,
I use Oracle 8.0.3 on Windows NT 4.0, 2 x Pentium 300MHz,
256MB ECC RAM, Adaptec AAA-133, 5 x 9GB UW SCSI(RAID 5),
2 x 4.5GB UW SCSI(mirror). DB block size 8k.
I have a large database. One of tables has ca. 55,000,000 rows,
second ca. 20,000,000 rows. The tables have daily increase
ca. 200,000 records. The data is transfer from ASCII with
comma separator to a temporary ORACLE table using SQL*LOAD,
than a query split it in two tables and the temporary table
is droped. Next day this same, etc.
When I created the database and transfer a lot of records
(ca. 48,000,000), the database crash.
part of or01ALRT.log
>>> Thread 1 advanced to log sequence 65533
>>> Current log# 1 seq# 65533 mem# 0: D:\ORANT\DATABASE\LOG1ORC1.ORA
>>> Thread 1 advanced to log sequence 65534
>>> Current log# 2 seq# 65534 mem# 0: D:\ORANT\DATABASE\LOG2ORC1.ORA
>>> Thread 1 cannot allocate new log, sequence 65535
>>> Checkpoint not complete
>>> Current log# 2 seq# 65534 mem# 0: D:\ORANT\DATABASE\LOG2ORC1.ORA
>>> Thread 1 advanced to log sequence 65535
>>> Current log# 1 seq# 65535 mem# 0: D:\ORANT\DATABASE\LOG1ORC1.ORA
>>> Thread 1 advanced to log sequence 65536
>>> Current log# 2 seq# 65536 mem# 0: D:\ORANT\DATABASE\LOG2ORC1.ORA
>>> Thread 1 cannot allocate new log, sequence 65537
>>> Checkpoint not complete
>>> Current log# 2 seq# 65536 mem# 0: D:\ORANT\DATABASE\LOG2ORC1.ORA
>>> kccrsz: denied expansion of controlfile section 9 by 65535 record(s)
>>> the number of records is already at maximum value (65535)
>>> krcpwnc: following controlfile record written over:
>>> RECID #1 Recno 1 Record timestamp
>>> 06/04/98 17:09:04
>>> Thread=1 Seq#=1 Link-Recid=0
>>> Low
>>> scn: 0x0000.00000001
>>> 06/04/98 17:07:56
>>> Next
>>> scn: 0x0000.0000006c
I know what the problem was - to many log sequences records
- to small redo log files.
I recreated the database with 100 times biger redo log files.
Now (today) the sequence log# is ca. 350, but after a few months I will
be
again close the maximum number of records.
I fund a info:
"There is a bug No - 485937"
but I don't have idea how solve the bug.
If somebody have any sugestions, please mail me back.
Thaks,
Andre
Received on Wed Jun 10 1998 - 00:00:00 CDT