Large data upgrade

This refers to this old post: We have a 3.2.1 system with loads of data and an upgrade to 4 takes a while. Busy getting DB sizes and metrics. Will update. Want to check my thinking. Optimal approach from what I can see is: First upgrade to 3.3.10 Upgrade to newest 4 (4.8.10) because these seems to have been some optimizations from 4.5+ regarding the migrations Is it advisable to go via the 3.3.10 step to save upgrade time? Alternatives are as noted in the previous thread: Put the DB on a diet first and then upgrade Spin up a new 4.8.10 "empty" instance and pump the data over
1 answers

Hey Jacques,

When I would migrate such old versions, I would always migrate first the last version of the major. You should make deployment packages for each major, so you can fix all warnings of deprecated functions etc.

I think it is smart to defragment your database first, to do so make a backup first and restore it to your database. That would make your database a little quicker.

Good luck!