Azure Blob Storage automatically partitions data based on blob name prefixes. Sequential or timestamp-prefixed names can create hot partition servers. This simulator helps you visualize how different naming conventions affect partition distribution and throughput.
⚙️ Configuration
0000001.json
500,000
📊 Results Summary
-
Estimated Partition Ranges
-
Max Blobs / Partition
-
Traffic on Hottest Partition
-
Peak Req/sec on Hottest
📈 Partition Server Distribution
Estimated blob count per partition range (based on name-prefix bucketing)
🌡️ Request Load Distribution
Estimated requests/sec per partition range
💡 Recommendations
📚 How Blob Naming Affects Partitioning
Azure Blob Storage Partition Facts:
• Blobs are partitioned by container + blob name
• Azure auto-splits partition ranges when traffic increases, but sequential prefixes limit this
• Single blob throughput: up to 60 MiB/s (500 req/s for smaller ops)
• Single partition range: up to 20,000 req/sec before auto-split is triggered
• Storage account: 20,000 req/sec ingress (region-dependent)
• Blobs are partitioned by container + blob name
• Azure auto-splits partition ranges when traffic increases, but sequential prefixes limit this
• Single blob throughput: up to 60 MiB/s (500 req/s for smaller ops)
• Single partition range: up to 20,000 req/sec before auto-split is triggered
• Storage account: 20,000 req/sec ingress (region-dependent)
| Naming Pattern | Partition Spread | Risk |
|---|---|---|
0001, 0002, ... | Poor — all names share prefix range | High — hot partition |
20260228/data_* | Moderate — grouped by date prefix | Today's range overloaded |
a3f_blobname | Excellent — hash distributes evenly | Low |
9876543210_* (reverse ticks) | Good — shifts hot range constantly | Low |
guid/blobname | Excellent — random distribution | Loses natural grouping |