I back up my server once a week, is that enough or should I do it daily?

weekly is way too little bro, you need to do it daily. and don't keep the backup local, go to the cloud — AWS, Google, Azure, Linode — they're all cheap. there's the 3-2-1 rule, we recommend it to everyone. have both local backup + cloud backup...

Backup automation tools: rsync (incremental), mysqldump (database), AWS backup service, Duplicati (free, encrypt capable), Backblaze B2 (1 cent/GB/month). Rotation strategy: Daily 7 days, Weekly 4 weeks, Monthly 12 months. E-commerce: ACID compliance, transaction logs, archiving binlogs. Restore testing should be done (monthly) — to verify the backup works. Off-site retention 30 days minimum, 1 year ideal.

Take a backup daily bro, what's the big deal write a script in a cron job, run it automatically. Go to the cloud too S3 is like 3 lira here negligible loss. If the server burns down even in a fire, if the data isn't in the cloud, then you're screwed...

Backup strategy planning: 1) Determine RPO (how many hours/days of data loss is tolerable), 2) Determine RTO (how fast restore is needed), 3) Retention policy (how long will you keep it), 4) Choose storage (local disk tape, cloud), 5) Encryption (for sensitive data) 6) Automation (cron scheduled tasks) 7) Monitoring (backup success rate), 8) Testing (quarterly restore test). Ideal e-commerce RPO: hourly (< 1 hour data loss), RTO: < 4 hours. Corporate standard: daily, but critical systems 4-6 hour incremental.

15 years of IT experience, and every single time I've faced a disaster it was because backups were neglected. Most individuals and small businesses back up weekly, but that's a bit risky. For e-commerce, you need at least one backup a day. Do a full backup monthly, and incremental for the rest. Go to the cloud, specifically use S3 with versioning enabled — so you can restore old versions.

Could you elaborate on that? The answer varies greatly by industry; there is no one-size-fits-all rule.

Payment information changes are never verified through the channel they came from. This is my opinion, I'm not claiming it's absolute truth.

Let me summarize the topic, since several different answers were given. Don't rely on a single measure; go layer by layer.

Just leaving this note, it might be useful.

I felt relieved reading this answer, so it's not just me. Trying to do this alone is the most expensive way.

Proven by experience.

If I understood correctly you're saying: When you try to change everything at once, nothing settles.

Correct me if I'm wrong.

There's one point I'm curious about. If permission and scope aren't in writing, don't start that test.

Let me speak from the other side; I'm on the supplier side. Just because everyone does it doesn't mean it's right.

Hope this helps.

Timely topic. If permission and scope aren't in writing, don't start that test.

When you try to change everything at once, nothing settles. If you post the result here, it will help others too.

I'm a small business, let me explain from my side. Hasty decisions become decisions you have to fix six months later.

Good luck with that.

Could you elaborate on that? tbh just because everyone does it doesnt mean its right.

If you post the result here it will help others too.

Do you think this works at any scale? If you get three different answers on a topic, the question was asked wrong.

Solutions that work at a small scale collapse when you grow; I learned this late.

Three different views emerged, they all complement each other. Most time waste accumulates in tasks waiting for approval.

Correct me if I'm wrong.

There is something to watch out for. Forgotten test environments are more often the entry point than live systems.

I agree with this. Hasty decisions become decisions you have to fix six months later.

That's all, sorry if I went on too long.