MICRO-OUTAGES · INTERMITTENT CONNECTIVITY

A two-second outage can disappear before your investigation even starts.

Continuously validate service availability and responsiveness so short interruptions, latency spikes and intermittent degradation remain visible after the network appears normal again.

If customers can notice it, your validation should be able to record it.

See intermittent degradation in Degust Explore Broadband

SHORT DOES NOT MEAN IRRELEVANT

A brief degradation across 6% of your customers is still a network event.

Short incidents may be difficult to reproduce manually, but frequent distributed validation can show when they occurred, how many locations were affected and whether they shared relevant characteristics. The 6% population shown here is illustrative.

affected

437 of 8,000 validation locations

≈ 5.5%

WHAT SHORT INCIDENTS LOOK LIKE

Intermittent problems leave different traces.

No single threshold defines every incident; the observable trace depends on the service and network context.

Latency spikeA normally stable connection suddenly becomes significantly slower for a short period.

Packet lossPackets disappear or require retransmission even though the connection remains nominally online.

ReachabilityA service or destination becomes unreachable briefly and then recovers.

Service degradationA real service becomes unusable even though infrastructure alarms may remain quiet.

Related insights

CAPTURE THE INCIDENT

Validate frequently enough to preserve what the customer experienced.

OBSERVE

Run continuous or scheduled validations from distributed locations.

DETECT

Identify short changes in reachability, latency or service outcome.

CORRELATE

Compare affected locations and supporting network evidence from the same time.

REVIEW

Look back at the incident after the network has returned to normal.

ONE CUSTOMER OR MANY?

Find out whether the interruption was isolated or shared.

Compare an affected population with similar locations that remained normal.

AFFECTED

DEGRADED

RegionNorthern Germany

TechnologyDOCSIS

CPEFRITZ!Box

FirmwareX.YY

CarrierCarrier B

RouteRoute B

COMPARISON

NORMAL

RegionNorthern Germany

TechnologyDOCSIS

CPEFRITZ!Box

FirmwareX.YY

CarrierCarrier A

RouteRoute A

Short incidents become operationally useful when affected locations can be compared with similar locations that remained normal.

REAL PRODUCT

Keep the incident visible after it has disappeared.

Review the time, affected population and supporting evidence from the incident window.

REAL DEGUST SCREENSHOTMICRO-OUTAGE PRODUCT VISUAL REQUIRED

01Timeline See when the degradation occurred.

02Affected locations Identify how broadly the event was visible.

03Evidence Compare network and service observations from the same period.

HISTORICAL INVESTIGATION

Go back to the moment the customer noticed the problem.

The network looks normal now. The complaint was yesterday evening.

Expected path

Observed path

Finding: investigation should use the network and service evidence from the incident window, not only the current state.

Degust retains validation results so operators can inspect what distributed locations experienced during a previous incident. Retention duration depends on the applicable subscription and configuration.

MICRO-OUTAGE QUESTIONS

Questions operators ask about intermittent connectivity.

Short incidents require frequent observation, context and historical evidence.

What is a micro-outage?

A micro-outage is a pragmatic term for a short interruption or degradation in connectivity or service availability. There is no single universal duration; the operational question is whether the event was noticeable and measurable.

How can an ISP detect a two-second Internet interruption?

Validation must run frequently enough from relevant distributed locations to sample the interruption. Recorded reachability, latency or service results can preserve evidence after connectivity recovers.

Why are intermittent Internet problems difficult to troubleshoot?

They may disappear before a manual test begins, affect only some locations or services and leave current-state monitoring looking normal. Time-aligned historical evidence is therefore important.

How can an ISP investigate a problem that has already disappeared?

Review validation and supporting network evidence from the incident window, then compare affected locations with similar locations that remained normal.

How can latency spikes affect customer experience?

A short rise in latency can delay interactions, disrupt real-time traffic or make applications feel unresponsive even when the connection remains online.

How can packet loss cause poor service even when the connection stays online?

Lost packets may require retransmission or arrive too late to be useful. Interactive, voice, video and gaming services can degrade without a complete loss of connectivity.

How can an ISP determine whether a short outage affected one customer or many?

Compare frequent distributed results across locations and group affected observations by region, technology, CPE, firmware, carrier, route or other relevant metadata.

Can continuous validation detect problems before many customers complain?

Frequent distributed validation can reveal shared degradation as it appears. Whether it precedes complaints depends on measurement frequency, coverage and the event itself.

How can short broadband incidents be correlated with CPE, firmware, route or carrier?

Attach available context to each result and compare the affected population with a similar normal group. Shared characteristics can focus investigation without proving root cause.

How can historical validation help investigate yesterday's network incident?

Historical results preserve what distributed locations observed during the incident window, allowing operators to examine the relevant state instead of relying only on today’s recovered network.

Don't let short incidents disappear without evidence.

See how continuous validation can preserve the customer experience during intermittent connectivity and micro-outages.

See Degust in action Explore Broadband