Query fingerprinting · Regression detection · Incident bundles
OroDb captures wait stats, fingerprints every query, detects regressions against a rolling baseline, and packages everything into a single incident ZIP — so you spend less time gathering data and more time fixing the problem.
No credit card · No agents · No schema changes on monitored databases
| Fingerprint | Normalized SQL | Status | CPU last | Baseline | Trend |
|---|---|---|---|---|---|
| a3f2c1b8 | SELECT ... FROM Orders WHERE CustomerId = @P | ⚠ +67% CpuMs | 207 ms | 124 ms | |
| 7c9d4e21 | UPDATE Inventory SET Qty = @P WHERE SKU = @P | ⚠ +31% Reads | 89 ms | 68 ms | |
| f1a8b3c9 | SELECT TOP @P ... FROM Products JOIN Category | Stable | 12 ms | 14 ms |
When things are on fire, you don't have time to manually pull wait stats, plan XMLs, alert history, and server config from six different places. OroDb packages all of it into a single ZIP — vendor-escalation-ready in seconds.
"The incident bundle is the feature that saves us the most time. What used to take 45 minutes to assemble for a support ticket is now one button."
From raw wait types to plain-English business impact — all signals in one place, with a structured workflow to action them.
Strips literals, normalises whitespace, hashes the SQL structure. Two parameterised queries doing the same logical work share one fingerprint — enabling true trend tracking across captures.
Regression DetectionReal-time and snapshot-based analysis, categorised into CPU, IO, Memory, Lock, and Other. Signal-vs-resource breakdown tells you instantly whether you have CPU pressure or data pressure.
Live DashboardHashes the structural shape of execution plans across snapshots. When a query's plan changes and performance drops simultaneously, OroDb flags it before your users notice.
Plan Shape HashNot a raw DMV dump. Each recommendation is automatically classified as HighConfidence, ReviewRequired, ConflictingDuplicate, or WriteHeavyRisk — so your team works the right ones first.
Workflow + ClassificationDefine thresholds per server: signal wait %, IO wait share, specific wait types. Evaluated on every snapshot with a 30-minute deduplication window. Critical alerts fire email automatically.
Email NotificationsAll alerts, regressions, plan changes, and new query fingerprints correlated chronologically. Zoom from 1 to 30 days to see exactly what changed and when — the DBA's post-mortem view.
Root-Cause CorrelationOroDb reads from standard SQL Server DMVs available since 2014. No extended events, no trace flags, no schema changes on your monitored databases.
* Azure SQL: server config section omitted from incident bundles (sys.configurations not available)
VIEW SERVER STATE. No write access to your monitored databases — ever.We'd rather you know upfront. OroDb is a focused investigation tool — not a full APM platform.
Installs in minutes. Point it at any SQL Server instance with a read-only login and it starts building your performance history immediately.
Add a monitored server with a read-only DMV login. Connection string encrypted at rest or stored in Azure Key Vault.
Manual on-demand or scheduled via cron expression. Each snapshot captures wait stats, top 20 queries, and writes fingerprint trend data points.
All alerts, regressions, plan changes, and new queries correlated chronologically. See exactly what changed and when.
Approve index recommendations, assign work, generate incident bundles, and watch baselines recovering over time.
OroDb adapts its interface to the user's role. DBAs see execution plans and DMV data. Executives see health status and business impact — no SQL required.
Full cross-tenant visibility and system configuration.
Manages servers, users, alert rules, and storage for their org.
Full access to snapshots, query analysis, plans, and the index workflow.
Health dashboard in plain English. No SQL, no plans, no raw data.
Every index recommendation moves through a structured process — classified automatically, assigned to a DBA, reviewed, and tracked to completion with a full comment trail.
DMV analysis identifies missing indexes with estimated impact. Auto-classified: HighConfidence, ReviewRequired, WriteHeavyRisk, or ConflictingDuplicate.
Route recommendations to specific DBA team members. Comment threads keep discussion in context — tied to the actual recommendation, not a Slack thread 3 days later.
Approve, reject, or defer with notes. The generated CREATE INDEX script is right there — test on non-production, then mark Applied when done.
After applying an index, watch the fingerprint trend chart. CPU and reads drop, the baseline recalibrates, the regression flag clears automatically over 3–7 snapshots.
Ready to stop guessing?
Connect a read-only login and see your first query fingerprints within 60 seconds. No credit card. No agents. No risk.