DMCDMC Bilgi Teknolojileri
Real-world Scenarios

The everyday issues a DBA actually faces

No marketing fluff — for each real DBA scenario, we show exactly which features of SentinelDB360 solve it, step by step.

Real DBA scenarios Step-by-step solution 4 engines, one roof KVKK & audit focused
Friday evening

Production blocking chain — pages won't load

Customer calls, the e-commerce site is unresponsive. No time to open SSMS and run sp_who2. With SentinelDB360, find the lead of the blocking chain at a glance and take a safe action.

  1. 1Go to the Blocking tab — the React Flow tree centers the head SPID; waiting sessions appear on side branches.
  2. 2Wait duration + lock type — each node shows wait_resource, wait_type, lock mode. Thresholds (warning/critical ms) drive color highlighting.
  3. 3Session input buffer — the last SQL executed by head SPID explains what it is doing.
  4. 4Kill SPID aksiyonu — from the UI with a confirmation dialog. System SPID (≤50) is auto-protected. Recorded in audit log.
Early warning

TDE certificate expires in 30 days

If a TDE certificate expires, data access is lost. SentinelDB360 alerts you weeks before certificates expire — no surprises.

  1. 1Go to the Compliance tab — TDE Status & Cert Expiry tab.
  2. 2get_tde_certificate_expiry for each certificate days_to_expiry calculated — 30/60/90 day thresholds.
  3. 3Audit raporu Auto-generated as Excel or PDF; sent to the compliance team.
  4. 4Renewal plan — new certificate is created and TDE is rotated; business continuity is preserved.
AI-powered

Query time grew 3.2× — why?

Same query took 200ms yesterday, 640ms today. Did the plan change? Parameter sniffing? Missing statistics? AI Insights answers the question.

  1. 1Go to the Insights tab — the changed query automatically appears at the top.
  2. 2get_query_store_regressions detects plan changes; old plan vs new plan comparison.
  3. 3AI commentary Generated by Azure OpenAI or local Ollama; English explanation: "The new plan was selected for parameter @order_id = NULL sniffing may be involved."
  4. 4Action — old plan is forced (force_query_plan) or OPTION (RECOMPILE) is recommended.
Backup validation

"We have backups" — can we actually restore?

Having backups doesn't mean restore is possible. Backup chain validation verifies the continuity of full / diff / log chains.

  1. 1— the changed query automatically appears at the top. — full/diff/log backups for the last 7 days.
  2. 2get_backup_chain_validation(days=7) chain-break check for each DB.
  3. 3Log shipping lag — log shipping delay to secondary (alert: > 1 hour → critical).
  4. 4Recovery model audit — alerts on unexpected FULL → SIMPLE change (early sign of log chain break).
Security

Failed login attempts for hours — brute force?

The failed_login_spike alert type counts failed logins in a 24-hour window. Warning if threshold is exceeded, critical at higher severity.

  1. 1Alert Center — click the failed_login_spike event.
  2. 2get_failed_logins(hours=24) with source IP, login name and error code list.
  3. 3Permission matrix shows the audited account's permissions; was sysadmin attempted?
  4. 4Webhook bildirim — Slack/Teams/PagerDuty'ye otomatik mesaj. SHA-256 dedup ile spam yok.
Pre-failover

Replica lag spiked — is failover safe?

Replica lag warns at KB or seconds thresholds. Before failover, see the state of Always On AG, PostgreSQL replication slot, MySQL group replication or Mongo replica set clearly.

  1. 1Go to the HA tab — engine-based topology view.
  2. 2SQL Server: get_alwayson_ag_detail log_send_queue, redo_queue, lag KB & sec.
  3. 3PostgreSQL: get_replication_slots retained WAL bytes; get_wal_archiver_status archive lag.
  4. 4MongoDB: get_oplog_window + replSetGetStatus per-member optime diff.
PostgreSQL

PostgreSQL: tables bloat, queries slow down

Autovacuum can't catch up, dead tuple ratio rises, queries read more pages. SentinelDB360 continuously tracks the vacuum/bloat profile.

  1. 1PostgreSQL → Vacuum sekmesi.
  2. 2get_vacuum_stats last_vacuum, last_autovacuum and dead_tup ratio for each table.
  3. 3get_bloat_stats table + index bloat % estimate; the most bloated candidates are listed.
  4. 4get_long_running_transactions idle-in-transaction sessions blocking autovacuum.
Compliance

Comprehensive audit report for the annual review

The independent audit team wants 30 days of audit log, the permission matrix and TDE/encryption status. Manual collection takes days; in Sentinel, one click.

  1. 1Go to the Reports tab.
  2. 2Excel/PDF export — 30-day audit_logs, permission_matrix, get_tde_status, get_audit_status.
  3. 3Compliance dashboard — CIS-style checks (sysadmin count, sa account, public role grants, default port, etc.).
  4. 4Based on the requested format the report is produced; data does not leave the organization (on-premise mode).

Test your scenarios in your own environment

14-day free trial — no credit card required. In the demo, we'll go through your specific scenarios together.

Request Demo Platform Capabilities