Dear Valued Customers,
We regret to inform you
that our US Linux Server (Earth) has caused various problems to you today, 28th
February 2007.
For your information,
at around 4.00am (GMT+8) this morning until around 8.00am (GMT+8), our Apache
server (HTTP) appeared up and down continuously. We found out that a client has
abused our server resource, whose script generated a lot of PHP processes until
our Apache server crashed. When the Apache server was crashed, it was rebooted
automatically after 3 minutes by our custom monitoring script installed on the
server. However, that particular client’s script again generated the same
amount of PHP processes to crash our Apache server, and Apache server was once
again rebooted by our script. This whole process lasted until our server was
rebooted manually at around 8.00am (GMT+8), and the client has now been temporarily
suspended for future investigation.
At the same time, our mail
server failed to deliver incoming emails to our clients’ mailboxes on the
server. All the incoming emails were queued in our incoming mail queue until
just now, which is around 8.30pm (GMT+8). Outgoing email however is fine. We
did not realize the problem as mail server appeared up and running until I
myself discovered that I have not been receiving email for the whole day, and
thus I checked the mail server log, realizing that I actually have a lot of
emails in the mail queue but at the same time my webmail received nothing. I then
immediately informed Tris to look into the problem, and I am relieved to tell
you that the mail problem has been fixed at around 8.30pm (GMT+8) and ALL your
email for today are in your respective mailboxes now. This problem could be
linked to the server abuse by client. We have reinstalled the mail server, Exim
and that fixed the mail server issue.
The above service port downtimes
(HTTPd & Mail) were not reported by Alertra & Hyperspin as both Alertra
& Hyperspin monitor our server by checking the ping response. During the
downtimes, our server responded to ping requests as it was still running. Hence,
Alertra & Hyperspin assumes that the server was running without problem at
all.
My system admins are
still working closely in diagnosing the server and monitoring its resource
usage. If you are experiencing any problems with the server, please do not
hesitate to submit a ticket to support@kkhost.net
or email to me at michaelfoo@kkhost.net.
We appreciate your cooperation!
Meanwhile, I would like
to apologize for any inconveniences caused.
Sincerely,
Michael Foo
KKHost.Net Founder & Chief Executive Officer
KK Host Network (1458953-U)