Scaling

Overview

Zektor.io supports vertical scaling — upgrading or changing the compute resources (CPU, RAM, storage) allocated to your database instance. You can scale your instance at any time from the dashboard.

Cluster Sizes

Each database engine offers multiple cluster sizes (tiers). Choose the right tier based on your workload:

PostgreSQL Tiers

Valkey Tiers

How to Scale

Scaling Up (Upgrading)

1. Navigate to your instance's Settings tab
2. Under Cluster Size, select a larger tier
3. Click Save Changes
4. The resize process will begin automatically

Scaling Down (Downgrading)

You can also scale down to a smaller tier to reduce costs:

1. Ensure your data fits within the smaller tier's storage limit
2. Navigate to your instance's Settings tab
3. Select a smaller tier
4. Click Save Changes

Note: Scaling down is only possible if your current data usage fits within the target tier's resources. If your database is using 8GB of storage, you cannot scale down to the Starter tier (5GB).

What Happens During a Resize

When you resize a PostgreSQL instance:

1. A new instance with the target cluster size is provisioned
2. Your data is migrated to the new instance
3. Connection endpoints are updated to point to the new instance
4. The old instance is decommissioned

Downtime: There is a brief period of downtime during the cutover, typically a few seconds. The exact duration depends on your database size.

Connections: Existing connections will be terminated during the cutover. Applications with connection pooling and retry logic will reconnect automatically.

In-flight queries: Any queries running during the cutover will fail. Ensure your application handles transient errors gracefully.

Billing: Tier changes are prorated to the day. You only pay for each tier for the days you used it.

Resizing Branches

Each branch can be independently sized. You don't need to match the cluster size of your main branch — this is useful for running cheaper development branches.

Storage Auto-Scaling

PostgreSQL instances with persistent storage support automatic disk scaling:

• Scale up: When disk usage exceeds 80%, storage is automatically increased by 5GB
• Scale down: When disk usage drops below 60%, storage can be reduced (deferred to off-peak hours to avoid oscillation)
• Limits: Auto-scaling respects the maximum storage limit for your tier

Additional storage beyond your tier's included amount is billed at €0.09/GB per month.

You can monitor your current storage usage in the Monitoring dashboard.

When to Scale

Signs You Need to Scale Up

Choosing the Right Tier

• Starter — Small projects, prototypes, development databases
• Standard — Production applications with moderate traffic
• Pro — Growing applications with significant query load
• Business — High-traffic production workloads
• Custom — Contact us for dedicated resources

Performance Optimization

Before scaling up, consider these optimizations that often solve performance issues for free:

PostgreSQL

1. Add indexes — Missing indexes are the #1 cause of slow queries

   -- Find tables with mostly sequential scans
   SELECT relname, seq_scan, idx_scan
   FROM pg_stat_user_tables
   WHERE seq_scan > idx_scan
   ORDER BY seq_scan DESC;

2. Use connection pooling — Reduce connection overhead with PgBouncer or application-level pooling

3. Optimize queries — Use EXPLAIN ANALYZE to identify slow operations

   EXPLAIN ANALYZE SELECT * FROM orders WHERE customer_id = 42;

4. VACUUM regularly — PostgreSQL needs VACUUM to reclaim space and update statistics

   VACUUM ANALYZE;

5. Check for table bloat — Large, frequently-updated tables can accumulate dead tuples

Valkey

1. Use appropriate data structures — Choose Sets, Sorted Sets, Hashes over individual keys
2. Set TTLs on keys — Prevent unbounded memory growth
3. Use pipelining — Batch commands to reduce round trips
4. Monitor eviction — If keys are being evicted, you need more memory

Need Larger Instances?

For deployments requiring more than our standard tier offerings — or for dedicated hardware, custom networking, or compliance requirements — contact us.

Next Steps

• Monitoring — Track resource usage and identify bottlenecks
• Connecting — Configure connection pooling
• Pricing — Understand costs for each tier
• Limits & Quotas — Storage and connection limits per tier