Redis Cloud generates a High Network Usage warning when your subscription or database approaches bandwidth or throughput limits. This can indicate saturation of monthly bandwidth on Essentials, configured ops/sec limits on Pro, or inefficient traffic patterns that risk throttling and latency spikes. This article explains what the warning means, how Essentials and Pro limits differ, how to determine whether you are bandwidth-bound or throughput-bound, and how to decide whether to Optimize, Increase throughput, or Upgrade plans. For step-by-step diagnostics, see Troubleshooting High Network Usage in Redis Cloud
What Triggers the Warning
You may see:
“High Network Usage”
“Bandwidth nearing subscription limit”
Throughput plateauing at a fixed ceiling
Elevated latency during peak traffic
This typically means one of the following:
Monthly bandwidth usage is approaching the subscription cap (Essentials).
Actual operations per second are reaching the plan’s maximum (Essentials) or configured throughput (Pro).
Large payloads or inefficient command patterns are generating excessive ingress or egress traffic.
Essentials vs Pro: How Limits Work
Essentials
Essentials plans enforce:
Fixed monthly bandwidth per plan tier
Maximum throughput per plan tier
Maximum connections per database
All databases share bandwidth and throughput quotas within a subscription. A single high-traffic database can impact others.
If bandwidth or throughput is exceeded, performance may be throttled and connections may be dropped. Increasing capacity requires moving to a higher tier or upgrading to Pro.
Pro
Pro allows:
Configurable throughput per database
Independent scaling of memory and throughput
No fixed Essentials-style monthly bandwidth cap. Network traffic is usage-based and billed according to the amount of data transferred.
Throughput is a sizing input that determines allocated CPU and network capacity. Actual performance depends on request size, command mix, replication, TLS, and client behavior.
Step 1: Determine What You’re Hitting
In the Redis Cloud Console:
Navigate to Database Configuration → Overview to view usage metrics, as these are reported at the database level rather than in the Subscription Overview.
Check monthly bandwidth usage and total ops/sec.
Database → Metrics
Review throughput trends, latency percentiles, and shard CPU.
Identify whether you are:
Bandwidth-bound (monthly usage near cap)
Throughput-bound (ops/sec at ceiling)
Experiencing both
Step 2: Choose the Correct Path
Optimize
Choose this first if:
Average payload size is large
Commands are complex or inefficient
Connection churn is high
Retry storms are amplifying traffic
Common improvements:
Use pipelining and batching
Replace KEYS with SCAN
Reduce JSON document size
Read only required fields
Fix connection pooling and backoff logic
See Troubleshooting High Network Usage in Redis Cloud for detailed diagnostics.
Increase Capacity Based on Plan
On Pro:
Database → Performance → Throughput
Increase configured ops/sec and monitor latency and CPU.
On Essentials:
Network capacity is fixed by plan tier.
If you are hitting network limits, you must upgrade to a higher Essentials plan to increase available bandwidth.
Upgrade to Pro
Upgrade when:
Network caps are repeatedly exceeded
One database affects others in the subscription
You need configurable throughput
Growth is sustained and predictable
Network-Intensive Workloads
Some workloads use minimal memory but generate heavy traffic:
Real-time messaging systems
Job queues
Large JSON reads
Traffic spikes in commerce or gaming
On Essentials, network bandwidth is tightly coupled to plan size and cannot be increased independently.
On Pro, throughput and infrastructure can be tuned more flexibly.
For sustained high traffic, Pro provides more predictable scaling.
When to Act Immediately
Take immediate action if:
Latency spikes correlate with throughput ceilings
Bandwidth exceeds 90 percent of monthly cap
Connections are being dropped
Other databases in the subscription are impacted
Summary
A High Network Usage warning means your Redis Cloud deployment is approaching bandwidth or throughput limits. The correct response depends on workload characteristics:
Optimize inefficient traffic patterns
Increase throughput where configurable
Upgrade plans when growth exceeds tier limits
Addressing the warning early prevents throttling, dropped connections, and subscription-wide instability.
0 comments
Please sign in to leave a comment.