Selecting an Event Database Solution
When selecting an event database storage system, consider the following factors over the life of the solution:
-
Online event storage requirements
-
Log ingestion rate
-
Large scale query and reporting performance requirements
-
3rd party application integration requirements
-
Resilience
-
Cost and complexity of deployment, and the ability of the organization to support it
The choice of event database storage can affect the cost, scalability and performance of the solution. Migrating between event database storage architectures is not always possible.
|
EventDB using NFS |
Single node ClickHouse |
Multi node ClickHouse |
External Elasticsearch[1] |
---|---|---|---|---|
Recommended EPS |
1 – 30K EPS |
0 - 20K EPS |
0 - 1M EPS |
0 - 1M EPS |
Large query performance |
Low - Medium |
Medium |
High |
High |
3rd party data access |
No |
No |
No |
Yes |
Resilience [2] |
High – depending on NFS |
Low |
High |
High |
Cost and Complexity |
Low |
Low |
Medium |
High |
Storage Efficiency |
High |
High |
High |
Low |
[1] Elasticsearch performance and resilience is dependent on the design of the external Elasticsearch cluster and the amount of resources provided to it.