High CPU usage on old database listener after data guard switchover
edited Mar 25, 2018 5:05AM in High Availability Data Guard, Sharding and Global Data Services (MOSC) 3 commentsAnswered
Hello,
after doing a data guard switchover (for Oracle 12.1.0.2 databases) from one Exadata machine A to Exadata machine B the Clients (Weblogic 12c) seem to work as expected. Connections to the new Primary database on machine B are established and the application is working fine.
But in the Background hundreds of connection attempts per second are done to the SCAN-listener on machine A (former primary, now the standby database), almost generating a denial of Service on machine A.
Has anyone seen this behaviour too ? Any hints to bugs or configuration issues (both on database or Client) side? I can't find anything Fitting to my problem in the knowledge base.
0