replication running very slow
Technical Note IRE2682Created 30-Dec-2004Applies To
DATABridge Host 4.2
affected structures allows duplicates (FXPYM in this case)
Symptoms
replication running very slowDataBridge up to 2 days behind in replicationlots of messages like this:
15:32:15 WARNING: Update for table fxpym had no matching row, changing it to an insert
- Keys: fxpym_actl_time = 8171159, fxpym_contractno = 315420053740, fxpym_input_rel = 17480
15:40:12 WARNING: Update for table benef had no matching row, changing it to an insert
Cause
Primary Key specified in GenFormat => customer was using a composite key instead of AA value as index
Because of that he wasn't detecting duplicate records which were updates. Instead of that DataBridge did a table search for every such record, after not finding a record to update it created a new record (insert). The extensive search made DataBridge run very inefficiently. Customer also ended up with much higher number of records then he should in reality have in the client's tables.