Timeouts and Latency
184 incident problems that show up as “Timeouts and Latency”.
먼저 읽을 가이드
추천 문제
All problems (184)
LINUX-026Swap activity stays high after memory spike is goneSwap activity stays high (Timeouts and Latency) is a hands-on troubleshooting drill. The incident is over, but swap churn remains elevated and keeps latency unpredictable for the rest of the day. Linux Service Operations needs to be checked by narrowing scope, recent change, a...LinuxAdvanced25 minProNETWORK-060SD-WAN policy prefers the worse pathThe controller has path metrics, but the application policy still chooses the higher-latency circuit because the SLA measurement feeding the decision is scoped incorrectly.NetworkAdvanced27 minProNETWORK-1029A QoS policy marks the right classes (Timeouts and Latency)A QoS policy marks the right classes (Timeouts and Latency) focuses on Network Services And Operations and asks the reader to isolate Timeouts and Latency in Cisco. 실무에서는 Timeouts and Latency 이슈를 볼 때 DNS, 라우팅, ACL/방화벽, 목적지 응답을 계층별로 잘라서 확인하면 장애 구간을 훨씬 빠르게 좁힐 수 있습니다.NetworkAdvanced28 minProLINUX-029Readonly remount happens only after storage latency burstThe filesystem looks healthy most of the time, but latency spikes trigger IO errors that remount the volume as readonly.LinuxAdvanced28 minProNETWORK-152An access switch voice VLAN is correct, but storm-control on multicast suppresses paging traffic only during announcementsThe network is healthy for normal calls, yet bursty multicast paging crosses a threshold the design did not anticipate.NetworkIntermediate15 minProLINUX-144nftables helper tracking is disabled globally, and active FTP works only until a maintenance reboot restores the stricter defaultThe fix looked stable, but it only lived in transient runtime state and did not survive policy reload.LinuxAdvanced16 minProLINUX-387A bonded uplink survives failover while one LLDP or MTU assumption remained on the old primary member only during a staged decommissionLink redundancy works and the operational metadata does not fail over with it. The service still works through the primary path, but one dependency only fails when the old component is finally drained away.LinuxAdvanced17 minProLINUX-159A host firewall marks packets for policy routing, but rp_filter on the return path still drops asymmetric replies after failoverForward policy is correct, yet kernel anti-spoofing logic rejects the recovery path.LinuxAdvanced17 minProLINUX-180A live firewall helper tweak fixes traffic until policy reload restores stricter persisted defaults during a failover rehearsalThe symptom disappears immediately and returns at the next state boundary. Normal traffic masked the issue until the standby or alternate path became active under rehearsal conditions.LinuxAdvanced17 minProK8S-393A multi-cluster failover record updates correctly while clients pinned to one HTTP/2 connection never rediscover the new backend set during a staged decommissionName resolution converges and long-lived streams keep the old world alive. The service still works through the primary path, but one dependency only fails when the old component is finally drained away.KubernetesAdvanced17 minProNETWORK-127A QoS policy references the correct class map, but after a subinterface migration it is attached in the wrong direction and voice jitter returnsConfiguration objects are present, yet the path where they matter no longer sees the service-policy.NetworkAdvanced17 minProNETWORK-252A QoS trust boundary shifts from phone to switchport while one access stack still rewrites DSCP at ingress during a failover rehearsalThe policy is consistent in design and one stack never received the new trust assumption. Normal traffic masked the issue until the standby or alternate path became active under rehearsal conditions.NetworkIntermediate17 minProLINUX-135A suspended virtual machine resumes with massive clock skew, but chrony makestep settings never allow the correction window to triggerTime drift is obvious, yet the recovery policy does not consider this magnitude eligible for immediate correction anymore.LinuxAdvanced17 minProK8S-137An external load balancer health check hits the NodePort before readiness gating is effective, and nodes enter rotation with no viable backendsInfrastructure sees open ports, but the application is not actually ready to serve traffic yet.KubernetesAdvanced17 minProLINUX-122An nftables set restores from disk without the interval flag and a whole CIDR block filter collapses into discrete entriesThe firewall loads successfully, but the restored semantics no longer represent the intended range-based policy.LinuxAdvanced17 minProNETWORK-357An SD-WAN overlay advertises the alternate circuit while local breakout policy still hairpins one SaaS prefix through the retired hub during a staged decommissionThe control-plane converged and the application policy path did not. The service still works through the primary path, but one dependency only fails when the old component is finally drained away.NetworkAdvanced17 minProLINUX-118cgroup v2 memory.high throttles the workers and the service looks slow instead of obviously brokenLatency climbs without hard OOM events because the workload is being pressured by the soft limit long before memory.max is reached.LinuxAdvanced17 minProLINUX-103Chrony sources are healthy but orphan mode keeps a node serving stale time after the upstream recoversNTP peers look reachable, yet one host continues advertising an old authority state because the failover mode never relinquished control cleanly.LinuxAdvanced17 minProK8S-126EndpointSlice hints keep steering traffic to a drained zone until the controller catches up, and clients see retries long after evacuationDrain activity succeeded operationally, yet topology-aware routing continues sending some traffic into the old fault domain.KubernetesAdvanced17 minProK8S-129Topology-aware routing and sticky sessions combine to create a hotspot on one zone after a scale eventNeither feature is individually wrong, but together they pin too much traffic to a shrinking endpoint subset.KubernetesAdvanced17 minPro