What could be the reasons of OPC-UA Collector taking a lot of time to parse the nodes?

We are using Greengrass V2 to ingest data from OPC-UA Server, and in a particular site we observed that the OPC-UA Collector takes a lot of time (more than 1-2 hours) to get started. We have 2 different data-sources defined in this case with same OPC-UA Server as source of data ingestion but different measurements for different sinks (Sitewise and Kinesis).

As per logs we suspect that it takes time to parse the nodes for both data sources considering both of them as separate Servers, and among them it prioritise the Sitewise first, where Kinesis Sink is more crucial for our use-case and as per our current understanding we can not provide the priorities among the data-sources. Even with feature of priority, it takes a huge time to fetch nodes.

What could be the reasons of OPC-UA Collector taking long time to get started, how to confirm the reason and how to solve this problem? Is there any optimisation we need to follow for Defining Data Source?

Connectivity and Configuration :

• Greengrass Machine is c6in.xlarge EC2 Machine on Cloud.
• The actual OPC-UA Server lies in remote machine, which is in a VPN link to the greengrass machine.
• Greengrass Version: 2.8.1
• SiteWiseEdgeCollectorOpcua Version: 2.1.3
• Node Structure: Object -> R1 -> R2 -> {device}_{measurement}, where R1 and R2 are directory nodes, and under R2 we have more than 4000 tags of name pattern {device}_{measurement}
• Data Source for Greengrass stream are defined as following:
  • Node ID for selection: /R1/R2/D*
  • Node Paths for Group Properties: /R1/R2/D*