Sre Glossary

Browse 31 sre terms defined in plain English, from the cultural dictionary of computing.

31 Sre Terms

Active-Active
Active-Active is a high-availability architecture pattern where two or more servers or data centers simultaneously handle production traffic, as opposed to...
Blameless Postmortem
A structured review conducted after an incident that focuses on understanding systemic causes and improving processes, explicitly avoiding individual blame —...
Bulkhead Pattern
Bulkhead Pattern is a software resilience design pattern inspired by the watertight compartments in a ship's hull that prevent a single breach from sinking the...
Chaos Budget
The amount of risk a team is willing to take with chaos experiments, analogous to error budgets. When reliability is high (error budget surplus), the chaos...
error budget
The acceptable amount of unreliability for a service over a given period, calculated as 1 minus the SLO. A service with a 99.9% availability SLO has a 0.1%...
Error Budget
The allowed amount of unreliability based on SLO targets — if your SLO is 99.9% availability, your error budget is 0.1% (about 43 minutes per month). When the...
Error Budget Culture
A culture that takes error budgets seriously as a way to balance reliability and change velocity. In healthy teams, error budget culture means service quality...
Failover
Failover is the automatic or manual process of switching to a redundant system, server, network, or component when the primary one fails, ensuring continuous...
Five Nines
A measure of system availability equal to 99.999%, allowing no more than 5 minutes and 15 seconds of downtime per year.
Golden Signals
Google SRE's four essential monitoring metrics: Latency (time to serve a request), Traffic (demand on the system), Errors (rate of failed requests), and...
Incident Commander
The person who coordinates the response during a production incident -- making decisions, assigning tasks, managing communication, and keeping the response...
Incident Management
The process of detecting, responding to, mitigating, and learning from production incidents. Includes on-call rotations, runbooks, war rooms, status pages,...
Load Shedding
The deliberate rejection of a portion of incoming requests when a system is under extreme load, preserving capacity to successfully serve the remaining traffic...
MTBF
Mean Time Between Failures. The average time a repairable system operates between successive failures, calculated as total uptime divided by the number of...
MTTF
Mean Time To Failure. The average time a non-repairable component or system operates before it fails for the first time.
MTTR
Mean Time To Recovery (or Repair). The average elapsed time from when a failure is detected to when the system is fully restored to normal operation.
On-Call
Being designated to respond to production incidents outside business hours. On-call rotation is the price of running things in production. The PagerDuty alert...
On-Call Rotation
A schedule where team members take turns being the primary responder for production incidents outside business hours. Good rotations include clear escalation...
RED Method
A monitoring methodology for request-driven services: Rate (requests per second), Errors (failed requests per second), and Duration (latency distribution)....
Retry Storm
A cascading failure where many clients simultaneously retry failed requests, overwhelming an already struggling service and preventing it from recovering —...
Site Reliability
The discipline of keeping services reliable through engineering approaches to operations, including automation, observability, incident response, and...
SLI
Service Level Indicator — a carefully chosen metric that quantitatively measures some aspect of a service's reliability, such as request latency, error rate,...
SLI/SLO/SLA
Three related reliability concepts: SLI (Service Level Indicator) is the metric measured (e.g., request latency p99). SLO (Service Level Objective) is the...
SLO
Service Level Objective — a target value or range for a service's reliability, expressed as a percentage (e.g., 99.9% availability). SLOs provide a...
tail latency
The response time experienced by the slowest requests — typically measured at the 99th or 99.9th percentile (p99/p999). In distributed systems, tail latency is...
Tail Latency
The response time experienced by the slowest requests — typically measured at p99 (99th percentile) or p99.9. In distributed systems, tail latency compounds:...
Thundering Herd
A concurrency problem where a large number of processes or clients are simultaneously woken up or triggered by a single event, causing a massive spike in...
toil
Repetitive, manual, automatable operational work that scales linearly with service size and provides no enduring value. Google's SRE framework defines toil...
Toil
Manual, repetitive, automatable operational work that scales linearly with service growth and provides no enduring value. A key SRE concept — Google's SRE book...
Toil
Manual, repetitive, automatable work required to run a production service. Google's SRE philosophy dictates that no more than 50% of an SRE's time should be...
Warm Standby
A redundancy approach where a secondary system is running and receiving data updates but is not serving live traffic, allowing it to be promoted to primary...

Related Topics