Understanding zero-downtime deployments for patches

So I’ve read the migration and upgrade guide pretty carefully and I still have a question.

Assuming you’re following some sort of blue green deployment model, it makes sense that if minor releases contain migrations you can only upgrade one minor release at a time to avoid downtime. But I am slightly unsure of patches and want to check my reasoning.

For example, let’s say you’re upgrading from 13.1.1 to 13.2.0 and you skip version 13.1.2 and 13.1.3 and 13.1.4.

(These versions are not exact. They are there for illustrative purposes only).

Well, I don’t see how you could skip the intervening three patches if all of them (unlikely) contain a migration. Hence, in my mind, the upgrade path would require doing all three patches first, and then the minor release. Is my reasoning correct?

(Staying with your non-exact version numbers).

GitLab maintains the three latest minor versions for security, so 13.2.0 might have been released before 13.1.3, meaning a lot of people will have upgraded from 13.1.2 til 13.2.0, naturally you can also do so after 13.1.3 and 13.1.4 have been released

In general you can trust that the patch releases won’t contain migrations.

Okay that makes sense why the documentation was not that concerned with this possibility. Makes sense, thanks for clarifying!