Symptom184 problems· 12 reviewed

Timeouts and Latency

184 incident problems that show up as “Timeouts and Latency”.

All problems (184)

NETWORK-099BFD over an encrypted WAN path stays up but the data plane breaksThe fast detection channel looks healthy, yet application traffic fails because larger encrypted packets are dropped midpath.NetworkAdvanced19 minProNETWORK-071BFD timers are too aggressive for WAN jitter and routing oscillates under light lossLinks are mostly usable, but routing keeps flapping because failure detection is tuned for a clean LAN rather than a noisy WAN path.NetworkAdvanced19 minProNETWORK-081BGP max-prefix protection drops the session during a route leak but only in one directionBoth peers are configured, yet only one side resets because the received leak exceeds the local protection threshold while the reverse direction stays below it.NetworkAdvanced19 minProNETWORK-102EVPN MAC mobility on the returning leaf uses a lower sequence and remote sites keep the stale locationThe server is back online, but east-west traffic still follows the old path because mobility ordering never tells peers to trust the new home.NetworkAdvanced19 minProLINUX-078nftables set refresh leaves stale IP entriesThe firewall update appears to run, but some stale members stay active because the atomic update path failed halfway and the old set was never swapped cleanly.LinuxAdvanced19 minProK8S-101StatefulSet ordinal restarts collide with the PodDisruptionBudget and quorum never reforms after node maintenanceOne replica keeps waiting for another to recover, but eviction protections prevent the exact restart sequence the quorum algorithm needs.KubernetesAdvanced19 minProNETWORK-088GRE keepalives pass but PMTUD blackholes the applicationTunnel control looks healthy, yet large application payloads fail because path MTU discovery signals never make it back through an intermediate filter.NetworkAdvanced20 minProNETWORK-077GRE tunnel MTU is too high and BGP over the tunnel flaps only during large updatesSmall keepalives succeed, but route churn triggers session instability because the encapsulated path cannot carry larger update packets cleanly.NetworkAdvanced20 minProCICD-096Rollback policy depends on ALB 5xx but the failure is only visible in gRPC status codesThe release is unhealthy, yet rollback never triggers because the monitored signal ignores application-layer failure semantics used by this service.CI/CDAdvanced20 minProNETWORK-095Route leaking between VRFs works in one direction but the return path missesForward reachability is proven, yet replies fail because the destination VRF only imports part of the leaked prefix set.NetworkAdvanced20 minProNETWORK-073QinQ outer VLAN mismatch isolates one tenant even though the provider handoff is upThe carrier circuit is healthy, but one tenant stays dark because the outer service tag expected by the handoff does not match on both edges.NetworkAdvanced21 minProNETWORK-067VRRP advertisements hit control-plane policing and the backup never sees master lossBoth routers are online, but failover stalls because the protection policy drops the very control traffic that should signal the state change.NetworkAdvanced21 minProNETWORK-066ECMP asymmetric return traffic breaks the stateful firewall even though both routers look healthyEvery routing table looks correct, yet sessions reset because the forward and return directions traverse different firewall state holders.NetworkAdvanced22 minProLINUX-065LVM snapshot reaches 100 percent and the application filesystem stallsThe origin volume still has free space, but writes slow or freeze because the snapshot copy-on-write area filled and the backup workflow never released it.LinuxAdvanced22 minProNETWORK-083MPLS L3VPN route is present in the core but import RT mismatch keeps the customer VRF emptyThe transport side looks correct, but the service remains dark because the destination VRF never imports the route target the provider is exporting.NetworkAdvanced22 minProNETWORK-058Controller telemetry fails although the management VRF can still ping the destinationBasic reachability tests look fine, but telemetry and API registration fail because the source interface or VRF binding used by the application differs from the test command path.NetworkAdvanced23 minProNETWORK-062MLAG peer link is healthy but the orphan VLAN is not allowed and ARP blackholes one rackThe chassis pair stays up, yet half the rack cannot resolve the gateway because the affected VLAN was never permitted across the peer path.NetworkAdvanced23 minProLINUX-054nftables hotfix disappearsA manual packet filter change solves the immediate outage, but the next reload or reboot brings back the old policy because the permanent source of truth was never updated.LinuxAdvanced23 minProNETWORK-085VXLAN VTEP peers are reachable but the overlay MAC table never learns one tenant segmentThe underlay looks healthy, yet one tenant still cannot communicate because the overlay learning path or segment mapping is broken for that VNI only.NetworkAdvanced23 minProNETWORK-063Route reflector strips a BGP community and upstream traffic engineering stops workingSessions remain established and routes are present, but the intended upstream policy no longer triggers because a critical community vanished in the reflection path.NetworkAdvanced24 minPro