Please, don't take my words as granted but talk to an actual lawyer. You'll probably even find a free session for startups somewhere in your city, at least in Europe.
https://ico.org.uk/for-organisations/guide-to-the-general-da...
https://ico.org.uk/global/contact-us/advice-service-for-smal...
What hav the Danish & Belgian regulators been doing lately?
The mere act of pulling all my database backups from glacier at once would cost enough to force me to just shut down my personal projects.
http://blog.quantum.com/backup-administrators-the-1-advice-t...
"The GDPR is open to interpretation, so we asked an EU Member State supervisory authority (CNIL in France) for clarification. CNIL confirmed that you’ll have one month to answer to a removal request, and that you don’t need to delete a backup set in order to remove an individual from it. Organizations will have to clearly explain to the data subject (using clear and plain language) that his or her personal data has been removed from production systems, but a backup copy may remain, but will expire after a certain amount of time (indicate the retention time in your communication with the data subject). Backups should only be used for restoring a technical environment, and data subject personal data should not be processed again after restore (and deleted again)."
Other opinions have concluded that you must keep an index of requested deletes in the face of backups, for instance.
You can still log accesses and aggregate them into statistics, just don't keep the IP addresses. You can still log IP addresses to detect DOS attacks or whatever, just delete the log when you don't need it anymore, after a day or so. There's no need to get backups from glacier, because you know there is no personal data in them.
You do need to have a documented and implemented backup retention policy and communicate this if you receive a request to delete a user's data.
Database backups are only a problem if you save them forever, though it sounds like you are. GDPR generally requires that you regularly archive, rotate out, and clean up old data.