The test results are based on a model in which the constant-bandwidth safety and security video traffic remained unaffected during a single node maintenance cycle, disk rebuild, or node failure.
We performed all tests as follows:
- With NIC failures to ensure a worst-case scenario for all sizing parameters.
- With drive failures in place in the cluster (for example, with PowerScale FlexProtect running) to ensure a worst-case scenario for all sizing parameters.
- With 10 GbE interfaces with no more than two SMB connections per interface. A 10 GbE interface can accommodate up to four Recorder connections at the maximum XProtect -supported values.
- With a per-camera bandwidth of 4 Mb/s.
PowerScale A3000
The following table provides bandwidth-sizing guidelines based on our test results.
Cluster | Security Center version | OneFS version | per node | Bandwidth MB/s (Mb/s) | Drive size | |
Per VM | Per PowerScale node | |||||
A3000 | 5.10 | 9.4 | 2 | 62.5 (500) | 125 (1000) | 12 TB |
Dell VxRail V670F
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.
Dell Technologies recommends:
- Keep some RAM (16GB) for the virtualization OS
- Do not run more than 6 VMs total
- Do not exceed VM designs above 1200 Mbps per server
The following table provides bandwidth-sizing guidelines based on our test results.
Server node | VMware version | VxRail version | Max. per node | Bandwidth MB/s (Mb/s) | Disk groups per node | No. of drives per disk group (Cache:Capacity) | |
Per Archiver | Per node | ||||||
VxRail V670F | 7.0.3 | 7.0.410 | 3 | 30 (240) | 120 (960) | 10 TB | 1:4 |
Configuration
Component | Details |
Storage | PowerScale A3000 |
OneFS | 9.4.0.10 |
Disk type | SATA |
Disk capacity | 12 TB |
Protection scheme | + 2d:1n |
Protocol | SMB |
Single node capacity | 200TB |
Component | Details | |
Storage | VxRail V670F | |
CPU | 2 X Intel Xeon Gold 6354 CPU @ 3.00GHz | |
Disk type | Type | SSD |
Cache | 800 GB mixed | |
Capacity | 3.84 TB | |
Protection scheme | Erasure Coding | |
NIC speed | 4X25 Gb | |
Storage controller | Dell HBA355i Fnt | |
Per node (%) | Max. CPU utilization | 28 |
Avg. CPU utilization | 24 | |
Max memory utilization | 29 | |
Per VM (%) | Max. CPU utilization | 34 |
Avg. CPU utilization | 30 | |
Max. memory utilization | 70 |
Component | Details |
Server | PowerEdge V670F |
CPU | Intel Xeon Gold 6354 3G, 18C/36T |
RAID policy | 5 |
vSAN protection scheme | Erasure coding |
Memory | 512 GB |
Component | Details |
Guest OS | Windows Server 2019 Standard |
vCPU | 8 |
Memory | 16 GB |
NIC type | VMXNET3 |
Video disk size | 2 TB |
Component | Details |
Camera type | Virtual camera |
Codec | H.264 |
Bitrate | 4 Mbps |
FPS | 30 |
Resolution | 1920 x 1080 |
Additional information from the test results are shown in the following graphics.
The average CPU utilization percentage was less than 65%, which is normal for a production environment during failure and recovery scenarios, such as disk rebuild.
The external network throughput rate displays the total amount of data that passed through the external network interfaces in the monitored cluster.
The disk throughput rate displays the total amount of data being read from and written to the disks in the cluster during a two disk rebuild.
The disk operations rate averages between 3 K/s and 5 K/s.