GCP Memorystore for Redis
You can enable this integration using the Connect a GCP Account option on the Monitor UI.
This integration has 32 metrics.
List of Alerts
| Alert | Description | Format |
|---|---|---|
| [GCP Memorystore Redis] Low UpTime | Uptime is less than 1 hour in a Memorystore Redis instance. | Prometheus |
| [GCP Memorystore Redis] High Memory Usage | High memory usage in a redis instance. | Prometheus |
| [GCP Memorystore Redis] High Number Of Connected Clients | Higher client connection count on a Memorystore Redis instance. | Prometheus |
| [GCP Memorystore Redis] Recurrent Rejected Connections | The number of recurring rejected connections on a Memorystore Redis instance. | Prometheus |
| [GCP Memorystore Redis] Low Cache Hit Ratio | The cache hit ratio is low in a Memorystore Redis instance. | Prometheus |
| [GCP Memorystore Redis] RDB Snapshot Error Status | The last RDB Snapshot status is Error. | Prometheus |
List of Dashboards
GCP Memorystore Redis
The dashboard provides information on the GCP Memorystore for Redis integration.
List of Metrics
| Metric name |
|---|
| gcp_redis_instance_clients_blocked |
| gcp_redis_instance_clients_connected |
| gcp_redis_instance_commands_calls |
| gcp_redis_instance_keyspace_avg_ttl |
| gcp_redis_instance_keyspace_keys |
| gcp_redis_instance_keyspace_keys_with_expiration |
| gcp_redis_instance_rdb_enabled |
| gcp_redis_instance_rdb_recovery_attempts_since_last_success |
| gcp_redis_instance_rdb_recovery_estimated_remaining_time |
| gcp_redis_instance_rdb_recovery_in_progress |
| gcp_redis_instance_rdb_snapshot_in_progress |
| gcp_redis_instance_rdb_snapshot_last_status |
| gcp_redis_instance_rdb_snapshot_last_success_age |
| gcp_redis_instance_rdb_snapshot_last_success_duration |
| gcp_redis_instance_rdb_snapshot_time_until_next_run |
| gcp_redis_instance_replication_master_repl_offset |
| gcp_redis_instance_replication_master_slaves_lag |
| gcp_redis_instance_replication_master_slaves_offset |
| gcp_redis_instance_replication_offset_diff |
| gcp_redis_instance_server_uptime |
| gcp_redis_instance_stats_cache_hit_ratio |
| gcp_redis_instance_stats_cpu_utilization |
| gcp_redis_instance_stats_evicted_keys |
| gcp_redis_instance_stats_expired_keys |
| gcp_redis_instance_stats_keyspace_hits |
| gcp_redis_instance_stats_keyspace_misses |
| gcp_redis_instance_stats_memory_maxmemory |
| gcp_redis_instance_stats_memory_usage |
| gcp_redis_instance_stats_network_traffic |
| gcp_redis_instance_stats_pubsub_channels |
| gcp_redis_instance_stats_pubsub_patterns |
| gcp_redis_instance_stats_reject_connections_count |
Monitoring and Troubleshooting Google Cloud Memorystore Redis
This document describes important metrics and queries that you can use to monitor and troubleshoot GCP Memorystore Redis.
The GCP metrics of type DELTA, which typically have the suffix _count in their name (but not always), return the cumulative value of 1 minute. They need to be divided by 60 to get the value per second.
Instance
Instance Status
Use the following query to check if your instance was rebooted recently:
gcp_redis_instance_server_uptime
A return value less than 3600 indicates that the instance was restarted in the last hour.
Memory Usage
Use the following query to get the memory usage of your instance:
100 * (gcp_redis_instance_stats_memory_usage / gcp_redis_instance_stats_memory_maxmemory)
CPU Seconds
Use the following query to get the CPU-seconds consumed by the redis server:
gcp_redis_instance_stats_cpu_utilization / 60
Keys / Cache
Number of Keys
Use the following query to get the number of keys in an instance:
gcp_redis_instance_keyspace_keys
Expired Keys
Use the following query to get the number of expired keys in your instance:
gcp_redis_instance_stats_expired_keys / 60
Evicted Keys
Use the following query to get the number of evicted keys in your instance:
gcp_redis_instance_stats_evicted_keys / 60
Cache Hit Ratio
Use the following query to get the cache hit ratio percentage:
gcp_redis_instance_stats_cache_hit_ratio
The returned value should be above 0.85. A ratio below 0.8 indicates that a high number of keys have expired/been evicted. Check for instance memory.
Network / Connections
Connected Clients
Use the following query to get the number of connected clients:
gcp_redis_instance_clients_connected
Blocked Clients
Use the following query to get the number of blocked clients:
gcp_redis_instance_clients_blocked
Rejected Connections
Use the following query to get the number of rejected connections per second:
gcp_redis_instance_stats_reject_connections_count / 60
A return value other than 0 means that maxclients limit has been reached. The default value of maxclients is 65000.
Replication
Pending Replication
Use the following query to get the number of Bytes pending to be replicated:
gcp_redis_instance_replication_offset_diff
Seconds Lagging
Run the following query to get the number of seconds that the replica is lagging behind the primary instance:
gcp_redis_instance_replication_master_slaves_lag
A value over 1 second indicates performance issues in the slave instances.
Pubsub
Pubsub channels
Use the following query to get the number of pubsub channels:
gcp_redis_instance_stats_pubsub_channels
Pubsub patterns
Use the following query to get the number of pubsub patterns:
gcp_redis_instance_stats_pubsub_patterns
RDB Snapshot
RDB Enabled
Use the following query to get the status of the snapshot mode:
gcp_redis_instance_rdb_enabled
If the instance is in production this metric should return 1.
Snapshot last status
Use the following query to get the status of the last snapshot:
gcp_redis_instance_rdb_snapshot_last_status
If the snapshot last status is other than 200 then there may be problems with snapshots.
Feedback
Was this page helpful?
Glad to hear it! Please tell us how we can improve.
Sorry to hear that. Please tell us how we can improve.