Our replicat ABENDED and would not start. The error states that the replicat was trying to modify the CURRENT_DIR column of the checkpoint table to NULL. This is by default a NOT NULL column. why would the replicat modify this column? The value of the CURRENT_DIR column for the replicat is the GG home directory.
2019-01-15T10:19:48.844-0600 ERROR OGG-00446 Oracle GoldenGate Delivery for Oracle, romhpt.prm: OCI Error ORA-01407: cannot update ("GG_ADMIN"."GGS_CHECKPOINT"."CURRENT_DIR") to NULL (status = 1407). Updating directory in checkpoint table gg_admin.GGS_CHECKPOINT, group ROMHPT, directory , key 3,414,715,040 (0x00000000cb886aa0), SQL <UPDATE gg_admin.GGS_CHECKPOINT SET current_dir = '' WHERE group_name = 'ROMHPT' AND group_key = 3414715040>.
2019-01-15T10:19:48.844-0600 ERROR OGG-01668 Oracle GoldenGate Delivery for Oracle, romhpt.prm: PROCESS ABENDING.
I was able to get the replicat started again after I modified the checkpoint table column CURRENT_DIR to NULLABLE. The replicat started and set the CURRENT_DIR to NULL and the replicat is running fine now.
1. has anyone encountered this issue?
2. what is the CURRENT_DIR column in the checkpoint table used for?
3. is there any repercussion to having the CURRENT_DIR column of the checkpoint table NULL?
4. is anyone else running with this configuration or issue?
I tested changing the CURRENT_DIR column back to NOT NULL and updating it with the previous value. The replicat gave the same error again and refused to start until it could update the CURRENT_DIR column to NULL again.
I am hoping someone can give me some information about this issue. I find nothing on the oracle support site matching this issue.
thanks for your consideration,
Jim