You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We has a full service outage from HE.net from approximately 15 December 2024 3:51am until 18 December 2024 00:29am in Amsterdam.
Our equipment continued running. Fibre links remained up, but upstream HE.net equipment failed. We lost all internet connectivity.
15 December 2024 03:53 - HE.net Outage starts.
15 December 2024 04:18 - OSM Ops Initial Email to HE.net NOC, asking if unplanned maintenance.
15 December 2024 04:24 - HE.net email response from HE.net NOC, not maintenance. Investigating issue. (Investigate why email receipt delayed to osmfoundation.org email)
15 December 2024 04:43 - OSM Ops phone HE.net for status update. Response paraphrase: "Outage in Amsterdam, we investigating. No Estimated time to recover. Equinix remote hands on-route"
15 December 2024 07:24 - OSM Ops ask for Estimated time to recover.
15 December 2024 11:12 - OSM Ops phone HE.net for status update. Response paraphrase: "No Estimated time to recover"
...
17 December 2024 12:21 - New Equinix Internet up and running. OpenStreetMap.org services restored.
18 December 2024 00:29 - HE.net confirm restoration of service.
TBC.
The text was updated successfully, but these errors were encountered:
Firefishy
changed the title
Produce post-mortem for HE.net outage - 15 December 2024
Produce post-mortem for HE.net outage in Amsterdam - 15 December 2024
Dec 18, 2024
Just as a Datapoint - no clue what HE said the cause was or what you gathered yourself. 2024-12-16 08:01 (CET) i had a look at a HE BGP Looking Glass for the route to planet.openstreetmap.org - at that time - 184.104.179.145. I have a screenshot of that output from core2.ams1.he.net showing 6 paths with an age of11 Days.
At the same time i did an "mtr" to the same address which was dropped on the first HE hop in my path in Düsseldorf, which i found astonishing. It may be that HE uses MPLS for forwarding and has ttl propagation off. Otherwise my guess would have been that some filtering on HEs routers went bad for that 184.104.176.0/21 on their border routers.
We has a full service outage from HE.net from approximately 15 December 2024 3:51am until 18 December 2024 00:29am in Amsterdam.
Our equipment continued running. Fibre links remained up, but upstream HE.net equipment failed. We lost all internet connectivity.
TBC.
The text was updated successfully, but these errors were encountered: