Incidents
The shared ticket inbox of the NOC shift.
P3 INC-1009 · Hard · reported by Luca Brunner (Service Desk)
Bern: WiFi slow in one area
-
Luca Brunner (Service Desk) Report 14:57:00
Some users in Bern complain that WiFi is unusable in their area since around 14:57, worse during the busy hours. Cable users don't notice anything. My first guess would be the application servers. Forwarded from the service desk, users keep calling. Can you have a look?
-
NOC-POC System 15:24:00
SLA breached: nobody solved this incident in time. The fault was repaired.
Post-mortem
- Root cause
- acc-brn-2 Gi1/0/24 ran at half duplex against a full-duplex peer: late collisions and load-dependent loss.
- Fault
- Duplex mismatch
- Right answer
- D: Set acc-brn-2 Gi1/0/24 back to auto-negotiation (duplex mismatch)
- Solved by
- Nobody: SLA breached
A duplex mismatch appeared on acc-brn-2 Gi1/0/24 at 14:56:59 (port forced to half duplex after a configuration change, peer at full duplex): errors on the half-duplex side and load-dependent packet loss, no outage.
Causal chain
- acc-brn-2 Gi1/0/24 ran at half duplex against a full-duplex peer: late collisions and load-dependent loss.
- What was noticed: slow, lossy connections in one area that got worse with load.
Why the other options were wrong
- Set acc-brn-2 Gi1/1/2 back to auto-negotiation (duplex mismatch) Wrong: it assumed acc-brn-2 had a duplex mismatch. Right place, wrong fault: acc-brn-2 had no such problem. The cause: acc-brn-2 Gi1/0/24 ran at half duplex against a full-duplex peer: late collisions and load-dependent loss.
- Send the cabling team: fibre cut between acc-brn-2 and dist-brn-2 Wrong: it assumed the cable acc-brn-2--dist-brn-2 was cut. acc-brn-2 Gi1/1/2 (between acc-brn-2 and dist-brn-2) was clean: normal light levels, no errors, no flaps. The cause: acc-brn-2 Gi1/0/24 ran at half duplex against a full-duplex peer: late collisions and load-dependent loss.
- Send the cabling team: fibre cut between ap-brn-2 and acc-brn-2 Wrong: it assumed the cable ap-brn-2--acc-brn-2 was cut. acc-brn-2 Gi1/0/24 (between ap-brn-2 and acc-brn-2) was clean: normal light levels, no errors, no flaps. The cause: acc-brn-2 Gi1/0/24 ran at half duplex against a full-duplex peer: late collisions and load-dependent loss.
Timeline
- 14:57:00 INC-1009 opened: Bern: WiFi slow in one area (Luca Brunner (Service Desk))
- 15:24:00 SLA breached; fault repaired
Inbox
9- P3 INC-1009 Bern: WiFi slow in one area SLA breached
- P3 INC-1008 Lugano: network alarm, no user impact so far SLA breached
- P2 INC-1007 Bern: network almost unusable SLA breached
- P4 INC-1006 Bern: hardware alarm, no user impact SLA breached
- P3 INC-1005 Bern: network alarm, no user impact so far SLA breached
- P2 INC-1004 Many failed logins on the network devices SLA breached
- P2 INC-1003 Bern: Intranet unavailable, devices not answering SLA breached
- P2 INC-1002 Bern: Intranet unavailable, devices not answering Resolved
- P3 INC-1001 Metrics and logs missing from some routers Resolved