About Recommendations and Considerations

The recommendations and considerations discussed here will help you better implement this architecture.

Although your requirements might differ from the architecture described here, you should use the following recommendations as a starting point.

Log Groups
Define multiple log groups to provide correct access permissions to different teams and avoid sharing sensitive data.
Cost Management
Logging Analytics service is charged on the volume of data in active and archival storage. To allow troubleshooting of day-to-day issues and obtain the benefits of anomaly detection, pattern detection and other machine learning capabilities, Oracle recommends using an active storage period of 90 days and moving logs older than 90 days to archival storage. Logs from stored archival records can then be quickly recalled on demand.

When designing a highly available application stack in the cloud, consider the following factors:

Performance
Query performance is based on the specified time range and the number of operations such as filters, group-by, and so on. For better query performance, Oracle recommends that you enrich logs with specific labels and fields at the time of ingestion.
Security and RBAC
Customize log source definitions to filter any personally identifiable information (PII) data, and enable geolocation enrichment.
Availability
Logging Analytics is a fully managed highly available SaaS service.
Deploy
The Oracle Cloud Infrastructure Logging Analytics app is available as a stack in Oracle Cloud Marketplace.