[Apologies for duplicate e-mails]
Dear Colleagues,
We have completed planned maintenance work on one of our border routers
that was causing network problems.
Unfortunately, the work took longer than expected and we encountered
some problems that affected our network connectivity and some of our
services.
We are pleased to announce that we have now resolved the problems and
everything is back to normal.
More details:
During the replacement of one of our routers, we discovered that MD5
passwords became corrupt, leading to broken peering sessions. By the
close of business on 4 November, we had re-established 90% of sessions.
On 5 November, we completed the planned maintenance and rebooted the
router. This reboot caused the RIPE NCC network to become unreachable by
that part of the Internet that preferred a BGP path to the router. We
detected and resolved the issue by 18.30 UTC on 6 November.
All peering sessions are now restored. Packet loss, which lead to the
original decision to replace the router, has disappeared.
The TTM and RIS services were affected by the network incident during
the weekend. After this was resolved, data collection started again and
due to the resilient design of the system, no data was lost.
We apologise for any convenience this may have caused. If you have any
further questions, please send an e-mail to <ops(a)ripe.net>.
Kind Regards
Ruud de Kooter
Operations Manager
RIPE NCC
Amsterdam