-1

Assume you're planning to have a PostgreSQL PaaS database server hosted in Azure/AWS/GCP. This PostrgeSQL server will contains GBs of spatial data (national land parcel polygons, address points etc.) stored in PostGIS enabled SQL tables. All tables have good spatial and non-spatial indexes. The database server SKU/config is powerful enough for heavy GIS data usage. The client computer (eg. staff laptops) connects to these cloud databases via a corporate VPN - in the office the speedtest.net results whilst using this VPN are: 60Mpbs down, 50Mpbs up, 16ms latency.

What is the minimum network requirement (latency, bandwidth etc.) for smooth rendering and querying of these PostGIS tables in QGIS/ArcGIS?

Any general guidelines would be useful here too. For example, is bandwidth more important than latency when rendering spatial data in GIS software?

It's hard to pose an exact question here as the use cases vary and the network requirement is less the more zoomed in you are and/or the fewer layers you have checked on (less data to show on map). I've struggled finding any online articles which cover this topic.

Theo F
  • 1,197
  • 1
  • 11
  • 18
  • If possible push the GIS layer to the cloud too, either through WMS or QGIS in the cloud – Ian Turton Apr 22 '22 at 14:41
  • @IanTurton I plan on serving some datasets from Azure Postgres using GeoServer (if the IT providers don't block it!), but most of the Azure Postgres (PostGIS) tables will be served straight into desktop installations of QGIS on client machines. It is the latter I'm trying to discover acceptable network performance figures for. – Theo F Apr 22 '22 at 14:44
  • then push your QGIS to the cloud if they absolutely must have access to the whole table, otherwise consider a WFS feed to go with the WMS – Ian Turton Apr 22 '22 at 19:29
  • @IanTurton sorry I’m unsure what ‘push your QGIS to the cloud’ means. Install QGIS cloud-side? It’s a desktop app though..? – Theo F Apr 23 '22 at 00:58

1 Answers1

1

Check the performance guidelines mentioned by Azure. The choice of using cloud for GIS software is optional and it is purely based on the application requirements. Kindly check the pricing and computation engine performance and validate with application requirements.

https://learn.microsoft.com/en-us/azure/azure-sql/virtual-machines/windows/performance-guidelines-best-practices-checklist

https://azure.microsoft.com/en-in/pricing/details/bandwidth/

Sairam Tadepalli
  • 1,563
  • 1
  • 3
  • 11