Google Memorystore Integration
Redis and Memcached instance metrics on GCP with auto-failover monitoring. Track memory pressure, eviction rates, and cache hit ratios with AI-powered anomaly detection.
How It Works
Create a GCP Service Account
Create a service account with the Monitoring Viewer and Redis Viewer roles. TigerOps uses the Cloud Monitoring API to collect Memorystore metrics without direct Redis or Memcached access.
Enable Cloud Monitoring API
Enable the Cloud Monitoring API and Memorystore for Redis API in your GCP project. TigerOps will collect memory, CPU, connection, and replication metrics from your instances.
Configure TigerOps Memorystore
Add your GCP project credentials and select the Memorystore instances to monitor. TigerOps auto-discovers Redis and Memcached instances across all regions in your project.
Set Memory and Failover Alerts
Configure memory utilization thresholds, eviction rate alerts, and failover event notifications. TigerOps fires predictive alerts before memory pressure triggers evictions or impacts application performance.
What You Get Out of the Box
Memory Utilization Monitoring
Track used memory, max memory, and memory fragmentation ratio per Memorystore instance. TigerOps alerts when memory utilization approaches your configured limit before eviction starts impacting cache performance.
Cache Hit Rate Analysis
Monitor keyspace hit and miss rates to measure cache effectiveness. TigerOps alerts on hit rate degradation and correlates drops with instance memory pressure, eviction events, or application deployment changes.
Eviction & Expiry Tracking
Track evicted keys per second and expired keys per second. TigerOps identifies when eviction rates spike and correlates them with memory pressure and the maxmemory-policy configured for your instance.
Connection Monitoring
Monitor connected client counts, blocked clients, and rejected connection events. TigerOps alerts when connection counts approach instance limits and helps identify connection leak patterns in applications.
Replication & Failover Health
For Memorystore instances with read replicas and high availability, TigerOps monitors replication bytes lag, replica offsets, and failover events. Alerts fire immediately when automatic failover is triggered.
Command Throughput & Latency
Track total commands executed per second, input/output bytes, and network throughput. TigerOps correlates command throughput spikes with application traffic changes and identifies unusually expensive Redis commands.
Memorystore Integration Setup
Configure TigerOps to monitor your GCP Memorystore instances with service account credentials.
# TigerOps Google Memorystore Integration
# Required IAM roles:
# roles/monitoring.viewer
# roles/redis.viewer
integrations:
gcp_memorystore:
project_id: "your-gcp-project-id"
credentials_file: "./tigerops-sa-key.json"
regions:
- us-central1
- us-east1
# Redis instances to monitor (empty = all)
redis_instances:
- prod-session-cache
- prod-api-cache
- prod-rate-limiter
# Memcached instances to monitor (empty = all)
memcached_instances:
- prod-object-cache
scrape_interval: 60s
metrics:
- redis.googleapis.com/stats/memory/usage_ratio
- redis.googleapis.com/stats/keyspace_hits
- redis.googleapis.com/stats/keyspace_misses
- redis.googleapis.com/stats/evicted_keys
- redis.googleapis.com/clients/connected
- redis.googleapis.com/replication/offset_diff
alerts:
memory_utilization_warning_percent: 80
memory_utilization_critical_percent: 90
eviction_rate_per_second: 100
cache_hit_rate_min_percent: 85
connected_clients_max: 1000
replication_lag_bytes: 1048576Common Questions
Does TigerOps support both Memorystore for Redis and Memorystore for Memcached?
Yes. TigerOps monitors both Memorystore for Redis (including Redis Cluster mode) and Memorystore for Memcached. Each engine has a dedicated metrics set with engine-specific dashboards in TigerOps.
How does TigerOps monitor Memorystore failover events?
TigerOps watches for role change events in Memorystore Redis HA instances via the Cloud Monitoring API. When a failover occurs and a replica becomes the new primary, TigerOps fires an alert with the timestamp, instance ID, and duration of the failover event.
Can TigerOps alert before Memorystore memory is fully exhausted?
Yes. TigerOps monitors memory utilization trends and fires a warning alert at your configured threshold (e.g., 80%) and a critical alert at a higher threshold (e.g., 90%). It also computes the time to full utilization based on the current growth rate.
Does TigerOps support Memorystore for Redis Cluster?
Yes. TigerOps monitors Redis Cluster instances and tracks per-shard memory utilization, command throughput, and hit rates. It identifies shards approaching memory limits and alerts before the cluster needs resharding.
Can TigerOps correlate Memorystore evictions with application performance degradation?
Yes. TigerOps AI correlates Memorystore eviction rate spikes with application latency increases in your upstream services. When cache evictions cause cache misses that result in expensive database queries, TigerOps surfaces the full chain of events in a single incident.
Catch Memorystore Memory Pressure Before It Causes Evictions
Memory utilization alerts, failover monitoring, and AI-powered cache hit rate analysis for Memorystore. Connect in minutes.