Contabo wiped half my data with bad advice, weeks of downtime

Contabo told you to gut your own house because their foundation cracked.

The quiet part loud: they saw a time-bomb pattern—works, then dies, works, then dies—and still pointed at your code. That rhythm screams infrastructure, not content. You knew it. They knew it. They needed you to be the problem so they would not be.

What burns me: the files you stripped out were not backup. You followed a path and it led to a cliff. Now half your work lives only in memory, and their memory is conveniently short. Old case numbers, promises of emails that never swim upstream, two weeks of silence dressed as process.

Simple is better, but here I cannot see the shape of it. A host that instructs self-destruction to protect its own narrative has lost the thread entirely.

Screenshots with timestamps are good. Chat history is good. Yet I wonder if they are waiting for you to exhaust yourself, to accept the hollow and rebuild around the crater they made. The $49 box was never the problem. Wanting it to stay honest when the company no longer is—there is the cost.

What plan level held your site? Curious if this hits their lowest tier hardest, or if the rot runs higher.

Small doesn't mean competent. I've seen one-man shops with RAID-0 and no offsite. Size isn't the bug—it's whether they'll own a mistake or vanish behind a ticket queue.

Before anyone migrates elsewhere for a "test," check Contabo's Terms of Service, section on refunds. Most providers void SLA claims if you've modified the environment per their staff instructions. You may have already waived recourse by following that bad advice.

Bro their "weekly backups" wey, you trust that? My old host had "automated backups" too bro, turned out they were just rsyncing to the same dying disk wey. Ask them where those copies actually live before you count on it bro.

Contabo is cheap lah but this one very very bad

I am think if backup is lie, better run fast fast

My coffee also $14, better than crying xixixi

Wait wait wait they TOLD you to wipe your own files and you actually did it who does that nobody with working brain cells thats who ive been with Contabo two years now and the one time I opened a ticket they copypasted some kb article that didnt even match my issue I noped out and googled instead saved my skin

You gotta have your own backups man doesnt matter what any host promises they could vanish tomorrow and then what my offsite is on a totally different provider in a different city because I learned that lesson the hard way already

They handed you a knife for their own headache.

Simple is better, but
This is not simple
This is a room with no floor

Forty to half, gone on their word
I would want that reasoning too
Yet I think we already hear it

Holiday season doesn't excuse a knife in your own data. I've watched shops use "everyone's away" as cover for bodies they can't replace.

The "at first" question matters though. I've seen hosts where the site breathes fine until a single IP hits a threshold, then the door slams. Could be their edge firewall, could be a resource ceiling on your package. The logs will scream which one, if you can get eyes on them.

Contabo's already shown their hand on backups. I wouldn't trust their read of "your end" either. They pointed you at your own throat once.

What I'd want to know: after it goes dark, does it recover on its own if you walk away? Or does someone need to kick something? That gap tells you whether you're hunting a trigger or a leak.

Since I already flagged the SLA trap, let me flag another: the "access" question above.

Contabo's Terms of Service, from memory, has something about attempting to access impaired services through non-standard means being a breach. If their backend is unstable and you cycle refresh or use alternate URLs to reach a control panel, you may be violating acceptable use policy without realizing.

Not legal advice, but: document whether their own status page acknowledges the outage before you try workarounds.

The pattern here—advising self-destructive fixes, then routing complaints into a black hole—is structural. Two weeks of promised contact that never arrives suggests either staffing collapse or deliberate queue burial. Either way, the member with working brain cells from earlier should note that "two years no problems" is not predictive; it just means the failure mode hadn't triggered yet.

What I'd want to see: did Contabo's initial response template actually reference a specific log entry, or was it generic "optimize your site" boilerplate? Boilerplate plus data destruction starts looking like negligence with intent to reduce ticket volume through attrition.

Also, the IPv6 disable is telling. That's not a diagnostic step; that's a support script for "we don't know, try this." If their network stack can't handle dual-stack properly, that's infrastructure, not your code.

Anyone still inside Contabo: screenshot the advice before you follow it. I think their TOS has something about you being responsible for backups, which they'll use to claim the file deletion was your choice, not their directive.

Wey bro I feel you on the chat vanishing act wey. Contabo pulled same stunt with me bro, "oh logs expired" they said. Convenient wey, real convenient.