Hello everyone. I work in the (mt) Media Temple Abuse Department. I
just wanted people to know that we were provided with very little
information to help us narrow down the "basic auth attempts" that were
causing Twitter problems and forced them to block a range of IPs
belonging to some web nodes. As of last night, at approximately 6:45pm
we were unable to find the source of the issue as it had most likely
subsided or stopped all together. These types of attacks are very
common, and as mentioned already, the quickest solution is to simply
ban the source. The attacks are generally an exploited server or site
that is running through a list of common user names and passwords,
then move on.
Generally, when these types of attacks occur, the source of the attack
is notified, and provided a snippet of logs that displays times the
attacks occur, source IP, useragent string, or any information of that
nature. Because we have a large client base that uses Twitter
legitimately, it makes it difficult to find a source based simply on
connections to twitter. We have created a tool that should help us
track down this type of problem in the future, if we know when it's
occurring. The checking that we have in place needs to be manually
implemented, so we will be making a conscious effort to be aware of
any notifications from Twitter moving forward.
A notification of our progress with the issue and request to remove
the blocked IP range was sent to multiple recipients at Twitter about
10:30 this morning. We have yet to hear anything in return regarding
this issue with them. We would like nothing more than to clear this up
as quickly as possible with them, people at (mt) use Twitter to. :)
M Handa