Home > Edge > Digital Cities Edge > Dell Validated Design for Urban Mobility with ISS - Design Guide > Sizing guidelines
This section provides a comprehensive overview of estimations for the required infrastructure for SecurOS installation.
The following table outlines the general sizing guidelines for each SecurOS module in a bare metal environment. More guidelines must be applied when high availability scenarios are being considered.
Max cameras (per server, running corresponding individual modules) | |||
| Low-capacity sample | Medium capacity sample | High-capacity sample |
SecurOS Auto (low) | 8 | 20 | 32 |
SecurOS Auto (high) | 4 | 10 | 16 |
SecurOS Cargo | 4 | 10 | 16 |
SecurOS Tracking Kit | 8 | 20 | 32 |
Sample specifications (per one server) | |||
RAM | 16 GB | 32 GB | 64 GB |
Number of CPUs | 1 | 2 | 2 |
Sample CPU1 Standard option | Intel Core i7-9700K (12M cache, 3.60 GHz) or better, 8 cores, 8 threads | 2 x Intel Xeon Silver 4210 (13.75 MB cache, 2.20 GHz) or better, 20 cores, 40 threads | 2 x Intel Xeon Gold 5218 (22 MB Cache, 2.30 GHz) or better, 32 Cores / 64 threads |
Sample CPU2 High-performance option | Intel Core i9-9900KS (16M cache, 4.00 GHz) or better, 8 cores, 16 threads | 2 x Intel Xeon Gold 6226 (19.25 MB Cache, 2.70 GHz) or better, 24 Cores / 48 threads | 2 x Intel Xeon Gold 5220S (24.75 MB Cache, 2.70 GHz) or better, 36 Cores / 72 threads |
OS Drive | 256 GB or more (SSD) | 480 GB or more (2 x SSD—RAID1) | 480 GB or more (2 x SSD—RAID1) |
Graphics | Standard SVGA | Standard SVGA | Standard SVGA |
Network | 1 Gb | 2 x 1 Gb | 4 x 1 Gb |
1 These CPU options offer standard performance and are compatible with SecurOS Auto (low speed), Cargo, Transit, and Tracking Kit modules. | |||
2 These CPU options offer high performance and are compatible with with SecurOS Auto (high speed). These scenarios require a higher performance per CPU core. | |||
GPU consideration: NVIDIA TESLA T4 GPU is used for the validation that is performed by Dell Technologies, set as a passthrough for the VM. Other GPU ISS recommends NVIDIA RTX4000, RTX2080 (or similar). Compute capability 6.1, 7.5, and 8.6. All detectors can use GPUs. If classification is used, GPU helps even more. Up to 30% more cameras per server can be used. For US and Middle East, the Recognition engine is neural network (NN)-based, and you must use the NN capture (NN algorithm). |
Based on the findings of the validation tests, SecurOS MCC, GIS, and Dispatch Center do not require high compute resources to handle the large number of cameras streams coming from remote sites and to perform the tasks associated with its capabilities. The table below depicts the typical specifications that are needed to run the specified modules.
Sample recommended specifications | |
Server | Dell PowerEdge R750 |
RAM | 16 GB |
CPU1 | 2 x Intel Xeon Silver 4309Y 2.8G, 8C/16T, 10.4GT/s, 12M Cache, Turbo, HT (105 W) DDR4-2666 |
OS Drive2 | 100 GB |
Network3 | 1 Gbps OR 10 Gbps depending on number of channels to be integrated |
1 These CPU options offer standard performance and are compatible with the modules above. The server is strictly used for administration purposes only and aligned with the recommended practice that a separate machine or workstation is used as a viewing client. If running SecurOS on a VM, the ratio between physical cores and vCPUs should be 1:2. Therefore, for an eight-physical core processor, 16 vCPUs should be allocated. | |
2 Hard disk and storage space requirement shall be considered accordingly if there is a need to store recording locally or on a centralized shared storage drive. Storage calculations can be done using the ISS web system design tool or the storage calculator of the IP camera vendor. For additional information about storage calculations, consult with the ISS presales team. | |
Network consideration: 3 Bandwidth calculation is crucial when large camera streams are propagated from remote sites. Network traffic from MCC server to any remote sites are calculated to avoid any network congestion on, both the remote sites and the MCC server. | |
GPU consideration: Only when video analytics is performed and to be considered on the same machine. Recommended GPU sizing and approved GPU list and information is provided in a separate document. | |
Supported VM platforms:
|