Table 8-39 shows the database status and recovery method in the event of an error during a reload operation. The recovery method depends on the options specified for execution of the database reorganization utility. If an error occurs during reload operation, check the specified options and the message output immediately before the error occurred, as shown in Table 8-39, and then recover the error.

Table 8-39 Database status and recovery method in the event of an error during reload operation

Option		Status/Recovery method	Message output immediately before error occurred
-l	-i		None	721	712 and 732	714 and 734	715	716
a	s	Status	Status before exe	Null*2	Null*2, *4	Table crtn cmpled
		Recovery	Re-exe	Re-exe reload*1	Re-exe reload*1
	c	Status	Status before exe	Null*2	Null*2, *4	Table crtn cmpled	Index not created	Index crtn cmpled
		Recovery	Re-exe	Re-exe reload*1	Re-exe reload*1		Create index	Unnecessary
p	s	Status	Status before exe	Null*2	Null*2, *4	Table crtn cmpled
		Recovery	Re-exe	Re-exe reload*1	Re-exe reload*1
	c	Status	Status before exe	Null*2	Null*2, *4	Table crtn cmpled	Index not created	Index crtn cmpled
		Recovery	Re-exe	Re-exe reload*1	Re-exe reload*1	Create index	Create index
n	s	Status	Not guaranteed	Null*2	Not guaranteed	Table crtn cmpled
		Recovery	Re-exe after recover database	Re-exe reload after recover database	Re-exe reload after recover database
	c	Status	Not guaranteed	Null*2	Not guaranteed	Table crtn cmpled	Index not created	Index crtn cmpled
		Recovery	Re-exe after recover database	Re-exe reload after recover database	Re-exe reload after recover database	Re-exe reload after recover database*3	Re-exe reload after recover database*3

Legend:

721: Row data deleted

712 and 732: Data reloading started

714 and 734: Data reloading completed

715: Index creation started

716: Index creation completed

: Not applicable

cmpled: completed

crtn: creation

exe: execution

Re-exe: Re-execution

recover: recovering

Note

Messages are abbreviated. For example, 712 means KFPL00712-I and 732 means KFPL00732-I.

^*1 If an error occurs during execution in the log acquisition mode or the pre-update log acquisition mode, the table data may become empty. If you execute an unload operation, null data is unloaded and the table data is lost. In such a case, re-execute the reload operation only.

If reorganization with the unload and lobunld statements specified and without the -j option results in re-execution of the reload operation, reorganize the table with the following procedure:

1. Specify only the unload statement to reload data into the LOB column structure base table.
2. Specify only the lobunld statement to reorganize the table.

^*2 Only the RDAREA indicated in a KFPL00721-I message becomes empty.

^*3 If the table storage RDAREA is different from the index storage RDAREA, an alternative method is to release the table storage RDAREA from shutdown status, re-initialize the index storage RDAREA, and then re-create the index (-k ixrc). For details, see the procedures for handling errors during batch index creation in the HiRDB Version 8 System Operation Guide.

^*4 When synchronization point specification is used, the utility stores as many rows of data as displayed in the KFPL00800-I message.