Guardian One

SQL Server

Best-practice summary: Validate each finding against workload, version, edition, dependencies, ownership, test evidence, approval, and rollback. Similar symptoms can have different causes.

Consolidated point-in-time findings from Guardian One assessment rules and supporting diagnostic evidence. Review recommendations with accountable service owners and qualified SQL specialists before making production changes.

Maintenance Suggestions

Best-practice summary: Define maintenance from recovery objectives, workload, data change, and available windows. Prove backup recoverability and integrity outcomes; do not schedule commands merely because they are available.

Review-only maintenance starting points based on the collected evidence. Guardian One does not execute these commands, create schedules, or replace customer monitoring. Confirm recovery objectives, maintenance windows, edition support, storage, permissions, and change approval first.

What	Why	Suggested timing	Example command	Risks and validation	Guidance summary

Application Database Checks

Best-practice summary: Preserve application contracts. Review schema, constraints, indexes, options, and query patterns with developers, then test correctness, concurrency, performance, and rollback before deployment.

Schema, integrity, configuration, indexing, and transaction-log checks for user databases. Findings are ordered by priority and should be validated against application behaviour before changes are made.

Persisted user-database inventory

This inventory provides assessment context even when no application-design exception was identified or persisted.

Database	State	Recovery model	Compatibility	Size	Last full backup
ERP_Core	ONLINE	FULL	150	286,720 MB	2026-08-13 02:00
ERP_Reporting	ONLINE	FULL	150	92,160 MB	2026-08-10 02:00

Showing all 1 application findings.

Attention	Group	Finding	Database	Details	Risk	Recommendation
High · P55	Index design	Review an index for a frequently joined ERP relationship.	ERP_Core		Referential operations may scan more data than necessary.	Validate the workload and proposed key order in a representative environment before change.

System Database Checks

Best-practice summary: Protect master, model, msdb, and TempDB according to their distinct recovery and operational roles. Test restoration and rebuild procedures before an incident.

Availability, access, recovery configuration, capacity, and backup evidence for SQL Server system databases.

Database	State	Accessible	Recovery	Compatibility	Size	Last full backup	Assessment	Recommended action
master	ONLINE	Yes	SIMPLE	150	512 MB	2026-08-13 01:00	No exception identified	Retain the evidence and continue scheduled review of availability, access, recovery configuration, and backups.
model	ONLINE	Yes	FULL	150	128 MB	2026-08-13 01:00	No exception identified	Retain the evidence and continue scheduled review of availability, access, recovery configuration, and backups.
msdb	ONLINE	Yes	SIMPLE	150	1,024 MB	2026-08-13 01:00	No exception identified	Retain the evidence and continue scheduled review of availability, access, recovery configuration, and backups.
tempdb	ONLINE	Yes	SIMPLE	150	8,192 MB	Not applicable	No exception identified	Retain the evidence and continue scheduled review of availability, access, recovery configuration, and backups.

Evidence Summary

• Databases: 6
• Configuration Checks: 3
• Security Principals: 3
• Performance Metrics: 14
• Backup Records: 5
• Consolidated Finding Rows: 3
• Maintenance Suggestions: 0

Optional First Responder Kit capabilities

Availability and execution status are assessed independently for each procedure. Configured Disabled means Guardian One detected the procedure but did not execute it. Missing, hidden, unauthorised, or configured-disabled procedures do not reduce native Guardian One coverage and are not interpreted as healthy or unhealthy SQL Server evidence.

Procedure	Collection status	Execution status	Toolkit database	Limitation
Not assessed. First Responder Kit capability evidence was not collected or persisted. Native Guardian One assessment evidence remains available; absence is not a pass.

Compliance Overview

Best-practice summary: Treat compliance as evidence of control operation, ownership, exceptions, and remediation—not a one-time score. Expire exceptions and retain approvals and validation records.

Evidence-led control overview for prioritisation and governance. This section does not constitute certification, formal attestation, or confirmation that all applicable organisational, contractual, or regulatory controls were tested.

Activity Overview

Best-practice summary: Activity evidence is a bounded sample. Correlate waits, blocking, grants, I/O, TempDB, and Query Store across representative periods before diagnosing a root cause.

Current requests and memory grants are a bounded point-in-time view. Current-request query text and user, host, and application identities are deliberately excluded.

Quantum advisories

Area	Observed evidence	Advisory
Current workload	0 user request(s); 0 blocked at collection time.	No blocking was sampled. Reassess during a representative busy period before concluding that blocking is absent.
Query memory	0 active grant(s); 0 waiting without a grant.	No grant wait was sampled. Use repeated assessments and Query Store evidence to validate workload behaviour.
Database file I/O	Worst busy-file average: 9.4 ms read; 3.8 ms write.	Correlate latency at or above 20 ms with waits, workload and storage telemetry; cumulative averages do not identify a single incident.
TempDB	37.5% allocated; version store 6.2% of data-file capacity.	Review allocation across several assessments, active versioning work, spills and autogrowth before resizing.
Historical query evidence	1 database status row(s); 1 review candidate(s).	Validate ranked candidates in their application context; do not force plans or change queries from aggregate evidence alone.

Current requests

Session / request	Database	State / command	Wait / blocker	Elapsed / CPU	Logical reads / I/O
No user requests were active at the point of collection.

Query memory grants

Session / request	Database	Requested	Granted	Used	Peak used	Wait
No query memory grants were active at the point of collection.

Database file I/O

Average latency is cumulative since SQL Server started or the file counters were reset. Physical paths are retained as repository evidence but omitted from this report.

Database	Logical file	Type	Reads	Average read	Writes	Average write	Size
ERP_Core	ERP_Core	ROWS	185,000	9.40 ms	47,000	3.80 ms	240.0 GB

TempDB allocation

This is a point-in-time allocation view and does not establish a growth trend.

Measure	Observed	Interpretation
Data-file capacity	8,192.0 MB	Total configured TempDB data-file size at collection time.
Allocated	37.5%	5,120.0 MB remained unallocated.
Version store	512.0 MB	Row versions retained at collection time.
Internal / user objects	1,024.0 / 256.0 MB	Workspace and user-object allocations.
Data files	8	Count only; file symmetry should be reviewed with the detailed file evidence.

Leading non-idle waits

Wait statistics are cumulative SQL Server evidence since the instance started or the counters were reset. Correlate them with current workload and query evidence.

Wait type	Area	Wait time	Share	Tasks	Average	Signal	Interpretation and next review
PAGEIOLATCH_SH	Data file I/O	82.0 sec	100.0%	180	455.6 ms	5.1%	Check storage latency, memory pressure, large scans, and indexes before attributing the wait to storage alone.

Query Store coverage

Database	State	Capture mode	Storage	Evidence window	Limitations
ERP_Core	READ_WRITE	AUTO	384.0 MB of 2,048 MB	2026-08-06 00:00 to 2026-08-13 17:00	Bounded retained evidence; Guardian One does not alter Query Store configuration.

Query Store review candidates

Candidates are ranked by total duration within the configured bounded window. The captured query excerpt keeps the report useful after the Query Store window changes.

Database	Query ID	Plan ID	Captured query	Executions	Average duration	Average CPU	Average reads	Safe next review
ERP_Core	1842	7	SELECT OrderId, CustomerId FROM dbo.Orders WHERE CustomerId = @CustomerId	920	1,180.5 ms	612.4 ms	18,400	Review the identified query and plan in its application context. Validate plan history, parameters, concurrency, and business criticality before tuning or forcing a plan.