NetBackup IT Analytics System Administrator Guide
- Introduction
- Preparing for updates
- Backing up and restoring data
- Monitoring NetBackup IT Analytics
- Accessing NetBackup IT Analytics reports with the REST API
- Defining NetBackup estimated tape capacity
- Automating host group management
- Categorize host operating systems by platform and version
- Bulk load utilities
- Automate NetBackup utilities
- Scheduling utilities to run automatically
- Attribute management
- Importing generic backup data
- Backup job overrides
- Managing host data collection
- System configuration in the Portal
- Custom parameters
- Performance profile schedule customization
- LDAP and SSO authentication for Portal access
- Change Oracle database user passwords
- Integrate with CyberArk
- Tuning NetBackup IT Analytics
- Working with log files
- Portal and data collector log files - reduce logging
- Data collector log file naming conventions
- Portal log files
- Defining report metrics
- SNMP trap alerting
- SSL certificate configuration
- Configure virtual hosts for portal and / or data collection SSL
- Keystore on the portal server
- Portal properties: Format and portal customizations
- Data retention periods for SDK database objects
- Data aggregation
- Troubleshooting
- Appendix A. Kerberos based proxy user's authentication in Oracle
- Appendix B. Configure TLS-enabled Oracle database on NetBackup IT Analytics Portal and data receiver
- Appendix C. NetBackup IT Analytics for NetBackup on Kubernetes and appliances
Performance Issues
Performance can be impacted by a number of issues. Use the following checklist to help you isolate problems.
Check the number of backup jobs that have been processed in the last 24-hour period.
Determine the level of database logging that has been enabled in
. If DBG level messages have been enabled, this can negatively impact performance. INFO and WARNING messages have negligible impact.Check if anything else is running on the server.
Note if performance suffers at specific times of the day and determine which processes are running during those times.
Verify the server's configuration: memory and CPU size
Check if Oracle has been tuned or if the default file (
) is in use.On Windows, by default, Oracle can use only 1 GB of memory. If the system has more memory available, use the /3GB option to tell windows to allocate 2 GB to the Oracle process.
Determine the top running processes--run
.What are the top running processes and what is the average CPU Utilization?
Is a lot of Virtual Memory being consumed?
Is there a high I/O Wait? (This implies a disk bottleneck.)
If oraclescdb is the top running process, using a lot of CPU resources:
Run:
This task will prompt for the process id (pid) and show the query that is running. If you run this several times over the course of 5-10 minutes and it returns the same query, you have a long-running query. Note that this will only work if you use process IDs associated with oraclescb processes.
Query the report database for long-running jobs.
Select from
table to see what reports have been running and for how long. This will help identify a report that someone may have inadvertently scheduled to run every five minutes.In
, use other sql queries:Run the above queries and capture the output to send to the Veritas Support.