Reliability & Recovery

How we keep your database running 		Page ← Home

← Back to Home

Database Down Right Now? Skip this page and go straight to Incident Response →

How We Keep Your Database Running

We're reliable. Not five-nines reliable. Not automatic-failover-in-seconds reliable. But your-database-stays-up-and-we-can-recover-from-disasters reliable.

Here's exactly what you get at $59-299/month from a bootstrapped service.

How We Think About Reliability

Three principles guide everything:

Isolation

Each customer gets dedicated or isolated infrastructure.

Dedicated tier ($119/mo):
Customer A → Contabo VPS #1
Customer B → Contabo VPS #2
Customer C → Contabo VPS #3

Shared tier ($59/mo):
Customers A, B, C → Same VPS
├─ Separate PostgreSQL databases
├─ Connection limits per customer
└─ Resource monitoring per database

Your slow query doesn't crash someone else's database. Dedicated tier gets hardware-level isolation. Shared tier gets database-level isolation.

Redundancy

Daily backups. Offsite storage. Weekly verification.

Your database
├─ Daily backup → Wasabi S3 (encrypted)
├─ WAL archives → Continuous
└─ Retention: 7-30 days (tier dependent)

We can restore to any point within your retention window. Backups tested weekly to verify they work.

Simplicity

Standard PostgreSQL. Proven tools. No proprietary systems.

Stack:
├─ PostgreSQL 16 (standard, not forked)
├─ pgBackRest (backup)
├─ pgBouncer (connection pooling)
└─ Standard pg_dump exports

Any PostgreSQL expert can understand our setup. You can export and migrate anytime. No lock-in.

What You Get

Uptime & Response

Backup & Recovery

Support

Data Portability

What We Don't Do

No Automatic Failover

We don't provide sub-minute automatic failover.

Why: Requires hot standby servers (2x cost), real-time replication, and constant monitoring. Enterprise solutions offering this start at $300-500/month.

Our approach: Fast manual recovery. Alerted → restore from backup → back up in 1-2 hours.

No 99.999% Uptime

We guarantee 99.9%, not 99.999%.

Why: Five nines means <5 minutes downtime per year. Requires redundant everything and ops teams on-call 24/7. Enterprise infrastructure starting at $500-1,000/month.

Our approach: 99.9% allows 43 minutes/month. Most months we hit 99.95%+ with zero unplanned downtime.

No Multi-Region

We don't run your database in multiple regions simultaneously.

Why: 3x infrastructure cost, complex replication, significant operational overhead. Makes sense for global apps, not most startups.

Our approach: Single region (you choose Germany or US), offsite backups in separate region.

No 24/7 Phone Support

We don't provide phone support.

Why: Solo founder. Email gives you written record of support interactions and allows for thoughtful responses.

Our approach: Email for support. Critical outages = IMMEDIATE response (automated alerts to my phone 24/7). Regular support = 4-hour SLA (business hours), but I typically respond within 30 minutes, 7 days/week.

Want Better Than Baseline?

Slack Support: +$29/month

(Included with Pro/Enterprise)

├─ Private channel with founder
├─ Real-time chat, 7 days/week
├─ Same 4-hour SLA (typically 30-min)
└─ Faster back-and-forth troubleshooting

Extended Backups: +$15/month

(Included with Pro/Enterprise)

├─ 30-day retention instead of 7-day
├─ More recovery points
└─ Longer point-in-time recovery window

Note: For true enterprise HA with automatic failover and five-nines uptime, expect $250-500/month minimum. We focus on reliable, affordable service for startups.

How We Verify It Works

Weekly Backup Verification (Automated)

Every Sunday:
├─ Restore random customer backup
├─ Verify PostgreSQL starts
├─ Run queries to confirm data integrity
├─ Measure restoration time
└─ Alert if anything fails

Monthly Recovery Drill

First Sunday each month:
├─ Full disaster recovery test
├─ Restore to brand new VPS
├─ Verify application connectivity
└─ Document timing and process

Quarterly Cross-Provider Test

Every 3 months:
├─ Restore to different hosting provider
├─ Confirm no vendor lock-in
├─ Test catastrophic failure recovery
└─ Update disaster recovery procedures

Results available on request. Email: jeremy@intentsolutions.io

Technical Stack

Infrastructure

Hosting:     Contabo VPS (Germany) / Hetzner (Germany/US)
OS:          Ubuntu 24.04 LTS
Storage:     NVMe SSD
Cost:        $10.35/month (we're transparent about this)

Database

PostgreSQL:  16.x (latest stable)
Pooling:     pgBouncer 1.21+
Config:      Production-proven defaults

Backup

Tool:        pgBackRest 2.51+
Method:      Daily full + continuous WAL
Storage:     Wasabi S3 (encrypted)
Encryption:  AES-256-CBC

Monitoring

Health:      Every 5 minutes
Metrics:     pg_stat_statements
Alerts:      Email + phone (critical)
Uptime:      Betterstack monitoring

Security

Firewall:    UFW (PostgreSQL + SSH only)
Intrusion:   fail2ban (auto-ban)
SSL/TLS:     Enforced
Updates:     Automatic security patches

Standard, proven tools. Any PostgreSQL DBA can understand this stack.

The Bottom Line

We're a $59-299/month database service run by one person. We're reliable for what we are. We're not pretending to be AWS.

You get:

You don't get:

If that works for your business, we're a good fit.

See Pricing | Security Documentation | Email Questions

Questions? jeremy@intentsolutions.io
Critical outages: IMMEDIATE (automated alerts 24/7)
Regular support: 4-hour SLA (business hours: M-F 9am-6pm ET)
Reality: I typically respond to everything within 30 minutes, 7 days/week.