Data Sharing longevity test results
The following are the results of a 50-user longevity test performed on the Cloudera Data Catalog and Cloudera Iceberg REST Catalog.
Test environment and dataset
The longevity test was conducted on an AWS Enterprise Data Lake cluster with the following dataset configuration:
| Configuration item | Value |
|---|---|
| Databases | 50 |
| Tables | 100 |
| Columns | 1000 |
| Snapshots | 0 |
| Data Shares | 1000 |
| Tables per Data Share | 5 |
| Users | 200 |
| Users per Data Share | 200 |
| Configuration item | Value |
|---|---|
| Databases | 250 |
| Tables | 20 |
| Columns | 1000 |
| Snapshots | 100 |
| Data Shares | 100 |
| Tables per Data Share | 15 |
| Users per Data Share | 501 |
1 Each user accessed one table from one database up to 50 Data
Shares.
The following table displays the longevity test results for 50 concurrent users on the AWS EDL cluster environment:
| Component | Users/Threads | Duration | Throughput | Total requests | Failed requests | Error % |
|---|---|---|---|---|---|---|
| Cloudera Data Catalog 2 | 50 | 72 hours | 2.03/s | 527373 | 62 | 0.01% |
| Cloudera Iceberg REST Catalog | 50 | 72 hours | 33.90/s | 8787271 | 0 | 0.00% |
2 The default Ranger heap size was
1 GB.
