All bites
The whole library, newest first. Filter by what you are here for, or pick a topic if you already know.
8664 bites
Page 27
Diagnosing a healthy p50 but breaching p99
One percent of requests are slow, hurting power users and fan-out calls; investigate GC, locks, contention, cold caches, retries.
Proposing availability and latency SLIs for an auth API
Availability as the ratio of successful valid requests; latency as the fraction served under a threshold; measure at the edge from the user's view.
What is an error budget?
Error budget is one minus the SLO, the allowed unreliability; it gates feature velocity versus reliability work.
Relationship between SLI, SLO, and SLA
SLI is the measured metric, SLO is the internal target for that metric, SLA is the external contract with consequences.
Symptom-based vs cause-based alerting
Symptom alerts fire on user-visible impact, cause alerts on internal conditions; page on symptoms to cut noise and catch unforeseen failures.
Head-based vs tail-based trace sampling
Head-based decides at trace start cheaply but blindly; tail-based waits for the full trace to keep errors and slow requests.
What is high-cardinality data in Prometheus?
Cardinality is the count of unique label combinations; each is a separate series; explosion blows up memory and query cost.
Black-box vs white-box monitoring for legacy apps
Black-box probes from outside, white-box reads internal state; start black-box for fast user-facing signal.
How does distributed tracing work?
Trace ID ties one request together, span IDs are individual operations, context propagates via headers.
Structured vs unstructured logging: why it matters
Unstructured logs are free-text lines hard to parse; structured logs are machine-readable key-value or JSON records; structure enables reliable querying…
Why use latency percentiles over the average?
Averages hide the tail and are skewed by outliers, so most users can suffer while the mean looks fine; use percentiles; prioritize p99 when tail users are high-value or fan-out…
What are the Four Golden Signals?
Name latency, traffic, errors, and saturation; explain each briefly; describe measuring latency as a distribution separating success from failure.
Explain the three pillars of observability
Metrics are cheap aggregated time series for detecting and alerting, logs are discrete event records for detail and context, traces follow one request across services to localize latency.
Error budget policy across dependent microservices?
Set per-service SLOs but anchor on user-facing journey SLOs, budget for dependency error via the multiplication of availabilities, attribute downstream-caused failures correctly, and use…
SRE vs traditional ops on a recurring alert?
Traditional ops repeatedly handles the alert manually and scales by adding people; SRE treats it as a bug, automates or eliminates the root cause, and scales sublinearly.
Risky launch with a near-empty error budget?
Contain blast radius via canary and feature flags, use progressive rollout tied to budget burn, negotiate explicit risk acceptance, add fast rollback.
What makes a blameless postmortem effective?
Cover summary, impact, timeline, root cause, action items; explain blameless means focusing on systemic causes so people report honestly.
How do you find and eliminate toil systematically?
Inventory and measure toil via time tracking or ticket analysis, prioritize by frequency times cost versus automation effort, automate highest-ROI items, measure success…
Error budget exhausted early: what now?
Invoke the error budget policy, shift focus from features to reliability, prioritize stability work, analyze what burned the budget.
How do you set SLOs for a service from scratch?
Identify critical user journeys, pick SLIs that reflect user happiness, measure current performance, set achievable targets with buffer, iterate.