Comment by OVH - Wednesday, 06 December 2017, 21:49 PM
On November 9, we encountered a big problem
on our optical network at RBX. The problem was related
to a software bug on the equipment we use
which caused the deletion of the configuration.
Since then we have updated the equipment on everything
our network. Also to prevent this type of bug from
never again causes a worry about our DCs, we have
decided to divide equipment clusters into 3
on the RBX website. So, if we ever have again
this bug, the configuration would only impact 30%
During the preparation of the maintenance that was to
start at 23:00, the configuration disappeared again
at 8:20 pm and all the links were down again. !!!!!
The database has been deleted while we are using
the latest software version. So there is another bug! http://travaux.ovh.net/?do=details&id=28835
We quickly returned the configuration. Some
links have returned, but not yet all. Currently
traffic passes Paris> GRA> AMS> RBX instead of
directly Paris> RBX. Hence the important latencies.
We look with Cisco to understand why all
the links are not UP while the configuration was
delivery to RBX.
We are not going to do the intervention tonight.
I want to understand why the configuration fades
and how can we do this maintenance without
again have the least concern about production!