Hello ppl,
as you probably have noticed, netlantis is down since a while. Netlantis had
critical performance problems and we decided to re-write some of the core
scripts to improve the DB update. While doing this we had some new ideas
about how to process the updates and much more has been re-written.
There are no new features for the users, just the core-system works on a
different way and all the BGP updates queuing problems will be gone.
"What's new then ?"
netlantis is running now since several weeks its own BGP daemon called
(guess) nbgpd. So far it works well and has pretty nice performance (able to
handle up to 500 full BGP sessions (tested up to 100)). "nbgpd" is dumping
updates into files (per peer, per timestamp) in its own (simple) binary
format.
The updates are then sent to "ndb" (netlantis DB manager) which is a fast
"routing table status file updater". This also runs since several weeks.
the last difficulty is to convert this "ndb" format for the SQL database (to
handle all the web/whois/telnet queries).
So far we wrote the MySQL (MyISAM format) data file properly.
"So? why isnt it up yet ????"
We still have a last part to finish, writing the MySQL (MyISAM format) index
file. Once this step completed, netlantis will come back up!
"What's the difference for the lambda user??"
Reliable informations! Until now, "global cisco bug upgrade worldwide in the
world" made netlantis to go booom due to the high load of bgp updates. This
will no more happen. We expect to have a delay to "real-time" of about 10 to
15 minutes, with few or many BGP updates, it wont matter!
"Well, I dont care about your internal stuff, when will it be back????"
We expect to bring netlantis back during the next week. We will inform you
then.
Best Regards,
Pascal Gloor
_ ______________ ___ _ ________________
/ |/ / __/_ __/ / / _ | / |/ /_ __/ _/ __/
/ / _/ / / / /__/ __ |/ / / / _/ /_\ \
/_/|_/___/ /_/ /____/_/ |_/_/|_/ /_/ /___/___/.org
... where all the routes meet
