Securely shuttling data to cloud tools & systems from on-premises sources, or traversing private networks. This checklist does not cover Personal Gateways, which generally are viewed as a Governance risk.
There are three different types of data gateways:
The current version of this checklist focuses upon On-premises data gateways.
[1] There are 3 types of on-premises data gateways:
1. Enterprise Gateway: For on premises data sources
2. Virtual Network Gateway: For traversing private networks
3. Personal Gateway: For small-scale, personal BI usage scenarios.
See the linked article in the checklist for more info.
[2] Spool storage will transiently store data on the Gateway machine. With an SSD this goes faster than a classical HDD. Check the linked article for more details.
[3] In smaller-scale situations, you might not need a cluster; however, avoiding the single point of failure plus load balancing is a huge advantage. It is generally worth it to set up a Gateway cluster.
[4] This is to ensure that queries from Development & Test workspaces do not interrupt Production workloads causing latency issues for business user queries.
[5] Load balancing ensures that if one Gateway is experiencing a high load, another will be used. You can adjust the thresholds for which this happens, yourself.
[6] Rarely, Gateway updates can cause issues with existing queries. It is thus a good practice to first update on your Development and Test Gateways to avoid situations where an update causes an outage in Production workloads.
Click one of the below for more checklists:
The below checklists are in progress.