MicrosoftMicrosoftThanks for the blog! You mentioned that 80 GB is “large” for an LOBs, but in reality, that’s quite small. Now imagine dealing with terabytes of data on our local PCs, which are so dependent on limited resources and network bandwidth. And on top of that, if you suggest using a VM to complement a PaaS Azure PostgreSQL Flexible Server, why add an IaaS solution to complete the job?
Do you have any optimized solutions or recommendations for this scenario? Thanks!
MicrosoftThank you for your feedback and for raising these points!
The 80 GB example was just a simplified scenario to illustrate the process using Azure PaaS services. The approach is not limited to that size and can be adapted for larger datasets or different environments, including on-premises or hybrid setups.
For terabyte-scale migrations with resource constraints, key optimizations include:
Parallel processing and chunking to handle large number of LOBs efficiently
Streaming techniques to reduce dependency on local resources
Network throughput tuning for faster transfers
Using a VM alongside PaaS is one option, but similar strategies can be applied without introducing IaaS if governance policies prefer avoiding it. The goal is flexibility—these methods work across infrastructures, not just Azure PaaS.
