Member
Shared, the cheapest one with the free domain. I just need my pottery shop to stay up, not run NASA.
Is there like a step by step somewhere for testing without breaking my live site? I don't have a second Hostinger account to try it on.
Shared plan you can just spin up a local VM with VirtualBox and import the cPanel backup there. Takes an hour, saves your business. Or ask Hostinger support if they have a staging environment, some do now.
I test my WordPress backups on a $5 Vultr instance I keep powered off. Fire it up, restore, check, destroy. Costs pennies.
Member
Not so bad if you think smart. My Vultr is $6, Hetzner storage box is 3.50 EUR, NAS is just electricity. Under $15 total. For data I cannot lose, this is cheap.
Before fire I paid one OVH server 15 EUR and thought I was safe. Wrong.
Member
I was in Strasbourg datacenter too. Not total loss for me because I had rsync to Hetzner Falkenstein running every 6 hours. Took 2 days to get DNS switched, but data was there.
OVH compensation was 1 month credit. Took 6 weeks to arrive. I moved everything to Hetzner after, never looked back.
Technically impossible metric, but I appreciate the energy. Packet loss is bounded at 100%. You likely mean observed zero reachability.
My PagerDuty data from that day: first alert fired at 00:47 CET, OVH status page updated at 04:13. 206 minute gap. Social media had photographs before any official channel. This is why I poll every 60 seconds now, and cross-reference with peer telemetry.
Acknowledged. Colloquial usage among peers. Meant observed 100% loss on all probed paths, which presents as undefined ratio when baseline is zero packets returned. The SLA math still holds: 10^4 seconds downtime, 10^1 percent credit per 10^3 seconds, theoretical maximum 10 percent monthly credit. Actual received: zero, physical disaster clause.
I have mtr logs if anyone wants packet archaeology.
Member
I'm new to this but my boss at ISP says OVH still has cheapest bandwidth in our region. We use them for CDN origin. After fire they added more Strasbourg redundancy but I don't know details.
I want to learn proper backup. Is BorgBackup good for beginners? I was looking at https://www.borgbackup.org but it's pretty dense.
Borg is excellent. I use it with Hetzner Storage Box via borgbase repo. Dedup is magic, my 200GB compresses to 60GB. Costs 9.50 EUR for 1TB.
That was a joke about people who think RAID is backup. It is not. RAID is availability, backup is recovery. Different religions.
ZFS send to another continent is real though. I do this to my brother's server in Toronto. Free except his electricity.
I run quarterly restore drills at work on AWS, but at home I got lazy. Fire thread made me check my personal stuff. Found my restic repository had broken snapshots for 4 months because I changed SSH key and never updated cron email.
Fixed now. Monitoring cron output is backup of backup.
Brother in suffering. What host was yours? Mine was some cPanel reseller, "automated" backup to same server. Useless when server gone.
Now I use rclone to Wasabi + local. Wasabi cheap cheap, $7/TB.
