The following is the minimum configuration required to support devices in a high-availability (HA) environment. Refer to Available channels per device for information about the supported number of channels.
| Type | Qty | vCPU | Memory | Disks |
|---|---|---|---|---|
RabbitMQ Node (ServAssure NXT only mode) | 2 | 2 | 4GB | 1x30GB OS 1x40GB /opt/arris |
| In an HA environment, 2 RabbitMQ nodes are required to achieve HA. | ||||
RabbitMQ Node (with customer queues enabled) | 2 | 4 | 16GB | 1x30GB OS Minimum 1x40GB /opt/arris Refer to RabbitMQ nodes for detailed disk requirements. |
| In an HA environment, 2 RabbitMQ nodes are required to achieve HA. | ||||
| ServAssure Collection Manager Node | n+1 | 4 | 28GB | 1x30GB OS 1x40GB /opt/arris |
| Notes: ServAssure Collection Manager is designed to scale out horizontally and operate in an active/active operational mode. Each ServAssure Collection Manager is capable of supporting 19 million channels. In order to achieve HA, one more manager node should be provisioned in addition to the quantity required to meet the desired scale. Enrichment mode provides a secondary data push to a message bus where the data is consumed by ServAssure NXT or 3rd party applications. The additional load requires more memory and CPU on the ServAssure Collection Manager. | ||||
| ServAssure Poller Node | n+1 | 2 | 14GB | 1x30GB OS 1x40GB /opt/arris |
| Notes: ServAssure Poller is designed to scale out horizontally and operate in an active/active operational mode. Each ServAssure Poller is capable of supporting 5.7 million channels. In order to achieve HA, one more poller should be provisioned in addition to the quantity required to meet the desired scale. | ||||
ServAssure TFTP Receiver (previously PNM) Node | n+1 (optional) | 2 | 14GB | 1x30GB OS 1x40GB /opt/arris |
Notes: A dedicated PNM node for TFTP data is optional, but if used one (1) dedicated poller is required in each polling region. This node is configured to process all PNM actions that require a TFTP server. This node does not support HA fail-over. An Ansible configuration change is required to configure this PNM node to fail over to another poller node. | ||||
ServAssure L2TP Receiver Node | n+1 (optional) | 4 | 8GB | 1x30GB OS 1x40GB /opt/arris |
Notes: ServAssure L2TP Receiver nodes are required to support Upstream Spectrum Analyzer in Static Pseudowire mode. At least one dedicated node for each polling region is required. While each L2TP node can manage an unlimited number of RPDs, each node can handle only 10 active L2TP sessions at once. To expand capacity for more simultaneous sessions, additional L2TP nodes can be added to the polling region. High availability is supported by deploying N+1 nodes. There is no impact on either the poller or manager nodes if L2TP nodes are added. The total device capacity of ServAssure Cache Node is slightly reduced for each L2TP node added to address the low latency processing requirements of streaming. | ||||