# Cutting P99 Latency 1000X During Connection Storms by Hardening ScyllaDB Admission Control

DevFeed: [Cutting P99 Latency 1000X During Connection Storms by Hardening ScyllaDB Admission Control](<https://devfeed.tech/articles/cutting-p99-latency-1000x-during-connection-storms-by-hardening-scylladb-admission-control-4869.md>)

Original publisher: [Read original article](<https://www.scylladb.com/2026/07/01/cutting-p99-during-connection-storms/>)

Author: Cynthia Dunlop

Published: 2026-07-01T13:35:06Z

Content type: article

Language: en

Sources: [ScyllaDB](<https://devfeed.tech/sources/scylladb.md>)

Topics: [Latency](<https://devfeed.tech/topics/latency.md>), [Databases](<https://devfeed.tech/topics/databases.md>), [Caching](<https://devfeed.tech/topics/caching.md>), [client](<https://devfeed.tech/topics/client.md>), [Processes](<https://devfeed.tech/topics/processes.md>), [Code](<https://devfeed.tech/topics/code.md>), [cpu](<https://devfeed.tech/topics/cpu.md>), [passwords](<https://devfeed.tech/topics/passwords.md>)

Tags: [2026-2](<https://devfeed.tech/tags/2026-2.md>), [caching](<https://devfeed.tech/tags/caching.md>), [code](<https://devfeed.tech/tags/code.md>), [cpu](<https://devfeed.tech/tags/cpu.md>), [customer](<https://devfeed.tech/tags/customer.md>), [database](<https://devfeed.tech/tags/database.md>), [deployment](<https://devfeed.tech/tags/deployment.md>), [engineering](<https://devfeed.tech/tags/engineering.md>), [latency](<https://devfeed.tech/tags/latency.md>), [performance](<https://devfeed.tech/tags/performance.md>), [process](<https://devfeed.tech/tags/process.md>), [processes](<https://devfeed.tech/tags/processes.md>), [scale](<https://devfeed.tech/tags/scale.md>), [systems](<https://devfeed.tech/tags/systems.md>)

## AI overview

ScyllaDB reduced P99 latency during connection storms by optimizing connection establishment through throttling, caching, password-hashing improvements, and service-level isolation. The article explains how connection surges, retries, task-queue pressure, and per-core connection handling can amplify load at scale.

## Source excerpt

ScyllaDB successfully mitigated performance-degrading connection storms by optimizing caching, throttling, and password hashing to achieve a 1000x reduction in tail latency