A glitch in the Matrix, or a hungry exploit?
Weird internet behaviour. What's going on?
Sûnnet Beskerming researchers observed an interesting deviation in global network traffic over the last 24 hours, particularly for South American, Asian, and Australian networks. Normally, global Internet traffic (as observed by the Internet Traffic Report) oscillates around nine per cent packet loss, with global response times of 138 ms, and the internally derived traffic index at around 79.
Sustained over the last 24 hours, the traffic index has dipped almost five per cent, packet loss has climbed to 11 per cent, and the global response time to almost 150 ms.
Normal spikes and dips as observed on the Internet Traffic Report show up as no more than three- or four-hour blocks of odd results before settling back into normalcy. This latest spike and dip has been sustained for at least 18 hours, with a rapid ramp up in the six hours prior to the peaks (and lows) being reached.
When the figures are considered against the seven-day average, and the 30-day average, the deviation appears to be quite significant and seems to mark a distinct event or set of events. When the reports for Asia, South America, and Australia are looked at in isolation, the three regions appear to be suffering from a related event, with similar patterns being observed in the data being put forward for those regions. Data for Europe and North America indicates that whatever is affecting the other regions, it isn't affecting Europe or North America. Independently sourced data at Keynote (using their Internet Health Report) indicates that there is nothing adversely impacting the US at this time.
Either these regions are experiencing the first stages of a global event, or they contain networks that are under a sustained attack for some specific reason.
So, what can be causing this problem? There appears to be nothing that is being reported by any of the usual agencies or news feeds, with SANS indicating a GREEN Threat level, and Symantec, McAfee, and the other major security software providers not indicating any new malicious software emergence.
Looking at the current Top 10 report from SANS, it appears that Port 5901 (used for VNC) is leading the charge for the top rating across all metrics (including a 20 per cent lead on the next port on the rising Trends chart). At the time of writing, the raw data for Port 5901 was showing disturbing results.
While there is spam, drive-by phishing attacks, and persistent worms attacking global networks, these have been ongoing attacks and should not be responsible for such a large change in such a short period of time by themselves.
If we consider port 5901 to be relevant to the reason behind the attacks, then we might have found a potential cause, and a potential target.
An exploit was added a couple of days ago to a number of security mailing lists, distribution sites, and other sources, which targets a remote code execution vulnerability in the AMX VNC ActiveX control. Since appearing on these sources it has spread to thousands of sites, and is guaranteed to have been seen by many, many people - some with malicious intent.
Although a remote code execution exploit is nothing special nowadays, this particular piece of code claims to achieve its goals without alerting the victim to the fact that they have just been successfully hacked.
Whether or not it is relevant to the real reason behind the observed response time and packet loss deviation will be seen over time. At the least, administrators and end users should keep a closer eye on their systems and networks over the next few days to see if this unknown problem is going to spread.
This article originally appeared at >Sūnnet Beskerming.
Sūnnet Beskerming is an independent Information Security firm operating from the antipodes. Specialising in the gap between threat emergence and vendor response, Sūnnet Beskerming provides global reach with a local touch.