Question about the SPOF in database level

Recently I’m proposing to use Sentry in the company. But I have a question about the “SPOF” in the document. I tried but failed to find any detailed explanation about it, so I have to post here to ask for help.

In https://docs.sentry.io/server/installation/#hardware , it notes that:

Sentry provides a number of mechanisms to scale its capacity out horizontally, however there is still a primary SPOF at the database level.

I’m a little bit confused about the SPOF here. The sentence seems that we can hardly scale in the database level of Sentry, but we can still do replication(master/slave or leader/follower) right? If that’s true, this doesn’t seem like a general SPOF issue. Does the SPOF here mean that: because we can hardly scale the PostgreSQL database horizontally, the capacity of the database will depend on the resource limit on the database server. When the database size is approaching the disk limit / memory limit, the single database server could be a single point?

Thanks!