The test results are based on a model in which the constant-bandwidth safety and security video traffic remained unaffected during select storage failure scenarios, such as disk rebuild, node failures, and failing network paths.
We performed all tests with disk drive failures, node failures, storage process failures, or NIC failures to ensure a worst-case scenario for all sizing parameters.
Tests focus on the rate which the ECS can reclaim space. The process to reclaim space once an object has been marked for deletion is called Garbage Collection (GC). The retention time that a video clip exists on storage is finite due to corporate policy, industry regulations, or governmental regulations. Typical retention times range from 3 to 12 months.
Dell recommends:
- Using vSAN or SSD-based configurations for GeoDrive cache disks.
- Calculating drive space requirements for local disk and ECS buckets based on the retention times used.
The following table provides bandwidth-sizing guidelines based on our test results.
ECS cluster | ECS version | GeoDrive version | No. of ECS nodes in test | Daily cluster GC reclamation rate (TB) | Per node-based max. bandwidth MB/s (Mb/s) | Max. aggregate cluster bandwidth based on tested cluster MB/s (Mb/s) |
EX3000 | 3.3.0.1 | 2.0.1.1 | 6 | 60 | 98 (784) | 588 (4704) |
EX300 | 3.4 | 2.0.2.2 | 5 | 19.5 | 38 (304) | 190 (1520) |
U4000 | 3.3 | 2.0.2.2 | 8 | 70 | 86 (688) | 688 (5504) |