|Applies To||RSA Key Manager Appliance 184.108.40.206|
RKM Appliances configured as multiple clusters joined in a group
|Issue||Apply error "ORA-01403: no data found" occurring on all clusters every 15 minutes|
The second and third clusters in the group initially show an Apply error "ORA-06575: Package or function HASH_VALIDATION is in an invalid state", and then subsequently only the other error "ORA-01403: no data found" repeats
Oracle Apply errors "ORA-01403: no data found" and "ORA-06575: Package or function HASH_VALIDATION is in an invalid state" started occurring soon after applying RKM Appliance patch 220.127.116.11
Both errors, ORA-01403 and ORA-06575, are seen on the /status/db page
|Cause||The error "ORA-06575: Package or function HASH_VALIDATION is in an invalid state" is caused by replication monitoring feature introduced in 18.104.22.168; this issue was resolved in RKM Appliance patch 22.214.171.124.|
The error "ORA-01403: no data found" is caused by replication monitoring feature introduced in 126.96.36.199. This error repeats on each cluster every 15 minutes which is due to the replication monitoring cronjob set by default (in 188.8.131.52) to run every 15 minutes.
|Resolution||If RKM Appliance patch 184.108.40.206 or a more recent patch is applied, the error "ORA-06575: Package or function HASH_VALIDATION is in an invalid state" will not occur as this issue was fixed in 220.127.116.11. However, if RKM Appliance patch 18.104.22.168 is currently applied and ORA-06575 error is being seen on the /status pages report, follow the steps below to clear this error.|
The problem with Apply errors "ORA-01403: no data found" re-occurring on regular basis (KMA-2446) has been fixed in RKM Appliance hotfix 22.214.171.124. The Apply errors can be cleared out using the procedure below.
*** It is highly recommended to contact RSA Customer Support for a review of /status/db reports before following the procedure below to clear out Apply errors, as the procedure will clear ALL Apply errors. ***
Follow the steps below to clear out the Apply errors:
1. Stop crond service on ALL appliances (after login as root):
service crond stop
2. Update permissions in Oracle db, carry out this step on each cluster (only on the primary node of each cluster):
su - oracle
sqlplus / as sysdba
SQL> GRANT EXECUTE ON DBMS_CRYPTO TO LOCAL;
3. Delete the Apply errors on each cluster (only on the primary node of each cluster) using the script "cs_delete_apply_errors.sh" (a part of /status pages):
(Note that this step may take a long time to complete depending on the number of Apply errors. For example, it took this script approximately 1 hour to clear around 30,000 such Apply errors. You can run this script concurrently on all clusters to save some time.)
4. Start crond service:
service crond start
5. Check /status/db report on all clusters to confirm that the Apply errors have been cleared.
6. If the number of Apply errors was very large (e.g., around 30,000 or so), it is recommended to restart Oracle on all clusters (to avoid potential Replication error ORA-23603). To do so, follow these steps:
- Login as root on each cluster
- To stop Oracle, run the command "service oracle stop"
- To start Oracle, run the command "service oracle start"
|Workaround||RKM Appliance patch 126.96.36.199 was applied|
|Notes||For more details about installing and generating /status/db report, see solution " How to check status or health of RKM Appliance for troubleshooting purposes? "|
|Legacy Article ID||a57132|