DDX for PostgreSQL — Service Status

Live availability for each public service, plus how fresh the ingested data is. There is no SLA on these services nor guarantee to any schedule or interval; these are observational only.

SERVICE UPTIME

Uptime over the last 15 minutes, 1 hour, and 24 hours. Cells without enough data yet show .

Service Now 15 min 1 hour 24 hours
HTTPS
MCP
NNTP
IMAP
POP3
Blended

— = no data for this window. 1 hour samples: up to 360 (TCP connect every 10 s). Latency for the most recent probe shown in the Now tooltip.

Last updated · source: /healthz

MCP TOOL UTILIZATION

Tool Calls (24h) Errors p50 15m p50 1h p50 24h p95 24h
Loading…

Frequency and latency over the last 24 hours, per MCP verb. Latency colour: green < 0.5 s · yellow 0.5–5 s · red > 5 s. Samples per window shown after the slash. Tools with zero traffic in the last 24 h are omitted.

REQUEST GEOGRAPHY (rolling 24h)

Where in the world the last 24 hours of requests came from, summed across every public surface (REST, GraphQL, MCP, git, mailing-list HTML, site). Country fill brightness is proportional to total request count; hover a country for the per-API breakdown.

Client IPs are looked up against the DB-IP Lite IP-to-Country database (CC-BY-4.0). Only the (country, API, minute) tuple is stored; IP addresses are never persisted. Aggregated from primary node (fra) traffic only — see /about for coverage notes.

SLOW QUERIES

Query Calls Mean Max Rows
Loading…

Top 10 statements by mean execution time from pg_stat_statements, excluding the introspection query itself and bare BEGIN/COMMIT/SET/SHOW. Hover the query cell for the full statement (truncated to 200 chars server-side). Same colour scale as MCP TOOL UTILIZATION: green < 0.5 s · yellow 0.5–5 s · red > 5 s.

HOST HEALTH

Per-host snapshot of CPU, memory, /data disk, disk I/O and network over the last 24 hours, sampled every 60 s on each node and averaged into 5-minute buckets. Sparkline colour switches to amber when the metric crosses its warning threshold and red when it crosses critical.

fra PRIMARY

CPU
MEM
DISK /data
DISK I/O
NET

gva REPLICA

CPU
MEM
DISK /data
DISK I/O
NET

Thresholds: CPU warn 60% / crit 85%, MEM warn 70% / crit 90%, DISK /data warn 85% / crit 95%, DISK I/O warn 20 MB/s / crit 80 MB/s, NET warn 5 MB/s / crit 50 MB/s. CPU and MEM are tighter on these shared 2 vCPU / 4 GB instances than they would be on dedicated hardware. Source: /api/v2/host-metrics.

DATA FRESHNESS

How recently each ingestion source produced new content. Ages are wall time since the latest record landed in PostgreSQL. There is no SLA on these intervals; they are observational.

Mailing Lists

InboxItemsLatest messageLast fetched
Loading…

Git Repositories

RepositoryCommitsSymbolsLatest commitLast fetched
Loading…

Other Sources

SourceKindLast fetchedNext sync
Loading…

INFRASTRUCTURE COSTS

Approximate monthly and annual cost to keep pg.ddx.io running. Figures shown in USD; Exoscale invoices arrive in EUR and are converted at €1 = $1.10. Source-of-truth: infra/docs/cost.txt in the infra repo (last reconciled 2026-05-25 after Bundle A: both hosts upsized to standard.large and 400 GB SSD).

Provider Service Monthly Annual
Exoscale fra Standard.Large (4 vCPU / 8 GB)$42.68$512.16
Exoscale fra block volume (400 GB SSD) $39.60$475.20
Exoscale gva Standard.Large (4 vCPU / 8 GB)$42.68$512.16
Exoscale gva block volume (400 GB SSD) $39.60$475.20
Exoscale Egress (within plan) $0 $0
Cloudflare DNS & CDN (Free plan) $0 $0
Cloudflare Load Balancer (Standard, ~50 k qry/mo)$5.50 $66.00
Cloudflare R2 storage (~600 GB w/ lifecycle) $9.00 $108.00
Cloudflare R2 Class A ops (writes, ~50 k/mo) $0.23 $2.76
Cloudflare R2 Class B ops (reads, ~10 k/mo) $0.01 $0.12
Cloudflare R2 egress $0 $0
Registrar pg.ddx.io domain ($12 / yr amortised)$1.00$12.00
AWS SES Outbound mail (alarms, ~1 k/mo) $0.10 $1.20
Total$180.40$2,164.80
Tip Jar [Donate]$0
Net (operator-funded)−$180.40−$2,164.80

Figures are pinned to the live IaC: Exoscale instance type and block volume sizes are read from infra/terranix/exoscale.nix; Cloudflare Load Balancer and DNS are read from infra/terranix/cloudflare.nix; R2 sizing is read from the breakdown in infra/docs/cost.txt. Re-reconciled against actual bills every quarter; if a number drifts, fix it there and the table follows on next site-deploy. The Tip Jar is a placeholder — the site does not currently accept donations.