AI Assistant
Contact UsDocsConsole
Alibaba Cloud Solution for Tair & Redis®
Your Redis Cluster
Deserves Better Than
Firefighting

Big Keys and Hot Keys silently degrade Redis performance, until they don't. DAS + SLS gives you the long-term visibility to govern them before the next incident.

See How It Works Contact Sales
  • Up to 10 Years

    max data retention

  • 10s

    Hot Key interval

  • $15

    per instance / month

  • 1 Click

    to activate

CORE CAPABILITIES
From Snapshots to Full Observability

Every capability answers a specific question your team has been asking — but couldn't with basic monitoring.

Key Lifecycle Tracking
Know exactly when every key appeared, changed, or disappeared

Four lifecycle states
existing, new, disappeared, new→gone — classified per key per query window

Boundary-tolerant logic
120s tolerance for Big Keys, 20s for Hot Keys, preventing false positives

Precise timestamps
new_at and disappeared_at columns for exact event reconstruction

Size & QPS Trend Analysis
Visualize key evolution over any time window you choose

max_size_kb / latest_size_kb
track Big Key memory footprint peaks and current state side by side

max_qps / avg_qps / latest_qps
distinguish flash-sale pulses from sustained bugs with a single query

PromQL trend charts
plug into SLS dashboards for a visual timeline of cluster key health

Proactive Alerting
Get notified before the incident, not during it

SQL-based alert rules
any query that returns rows triggers a notification for full alert-condition flexibility

Three built-in scenarios
Big Key count threshold, Hot Key QPS spike >10,000, new key sudden appearance

SLS Alert Center
configure notification channels, escalation policies, and on-call schedules

Governance Validation
Prove that your cleanup actually worked, with data

Before/after comparison
query the same key over two time windows to measure fragmentation reduction

Reappearance detection
a "disappeared" key should not come back; set alerts if the same key re-enters the Top Keys list

Flexible period-over-period analysis
compare QPS trends with custom periods (Daily/Weekly/Monthly) to validate cache/splitting decisions by benchmarking against flexible historical baselines

SLS MetricStore Schema

Fields delivered to redis_top_key_log.prom


__labels__instance_id, node_id, key, key_type, topkey_type


__value__Size in bytes (Big Key) · QPS lower bound (Hot Key)


__time_nano__Millisecond-precision collection timestamp


__name__bigKeyItemCnt or hotKeyQpsLowerBound


Retention & Requirements

What you need to get started


Cloud-managed Tair/Redis® instance
DAS O&M Service enabled (paid feature)
SLS auto-configured on DAS activation


Default retention
30 days



Maximum retention
10 years


WHY UPGRADE
Basic Monitoring vs. DAS + SLS

The built-in console gives you a window into right now. DAS + SLS gives you the full story.

CAPABILITY REDIS CONSOLE (Basic Monitoring) DAS + SLS SOLUTION
View current Top Keys
Historical trend analysis
PromQL + SQL
Key lifecycle classification
new / disappeared / existing
Big Key size changes over time
max_size_kb / latest_size_kb
Hot Key QPS trend analysis
max / avg / latest QPS
Custom time range queries
Parameterized SQL
Long-term data retention
Up to 10 years
Proactive alert notifications
SLS Alert Center
Governance validation
Before/after comparison
Dashboard & scheduled reports
SLS Dashboard
Basic monitoring reaches the accident scene. DAS + SLS reconstructs the full truth — and predicts the next one. Read the Guide →

START TODAY
Stop reacting.
Start governing.

Enable DAS O&M Service with one click and get full lifecycle visibility in minutes, not weeks.

Download the Guide