Hi everyone,
A little while ago our friends at Phyber - http://www.phyber.com/ - racked a
few more servers for us and installed RHEL 6 on them. Over the last months
I've been moving services to the newer boxes.
Today I moved the web frontend (previously just Varnish on one box) to two of
the new servers (keepalived/stud/haproxy/varnish)[1] so there's appropriate
redundancy at that layer. Let me know if you notice any problems. One of the
maybe-user-visible changes is that the list of SSL ciphers for the manage site
has been shortened quite a bit.
The actual website is still just running on one of the older servers; the old
RRD based graphs make it a bit more hassle to move that to have redundancy, but
it's next on my list.
http://www.pool.ntp.org/
https://manage.ntppool.org/
Ask
[1] Keepalived to move the public facing IP around. Stud to terminate SSL for
the manage site (and the main pool site in the future). HAProxy because it can
get the 'real IP' from stud easily and because some traffic shouldn't go
through Varnish. Varnish because the server running the site is real slow and
sometimes the site gets thousands of requests in a short timeframe (seconds to
a minute).
--
http://askask.com/
_______________________________________________
pool mailing list
pool@lists.ntp.org
http://lists.ntp.org/listinfo/pool
