Distributed denial of service (DDoS) attacks are proliferating around the globe, impacting big brands and organisations from every sector.
In fact, new research from American technology company Neustar, shows that 73% of more than 1,000 IT professionals from the technology, financial, retail and government sectors suffered a DDoS attack last year, with many of them hit repeatedly, some six times or more.
On top of this, more than half reported some form of theft afterwards, including loss of intellectual property, customer data and money, the research finds.
What is a DDoS?
Vector Communications, the Auckland based company that specialises in building data networks, says DDoS occurs when hundreds, possibly thousands, of compromised computers, start flooding your IT system with an overwhelming volume of data every second. The likely source of this traffic is a botnet, which can be used to generate the massive amount of packets (data) needed to create a successful DDoS attack.
The type of attack may vary, from direct requests to your nodes from infected computers on the botnet to more sophisticated indirect methods like DNS spoofing, where a botnet sends queries to internet name servers while pretending to be one of the servers of the business (with a spoof IP address).
The result is a multiplier, where the hundreds or thousands of compromised computers are able to use name servers to generate a huge volume of data directed at your node.
Whatever form the attack takes, the end result is the same - a huge number of packets will be headed for the business and they’ll be coming from many different sources. That’s the distributed part of the attack, Vector says.
The ‘denial of service’ is when hundreds of thousands of packets per second are concentrated on either the internet connection or the connections of upstream ISP and start to aggregate. At this point something is going to fail and it will happen one of two ways, Vector says.
First, the total DDoS traffic will saturate or exceed the size of the organisation's internet link. While a business has likely planned their bandwidth needs based on reasonable expectations of use, what is happening now isn’t reasonable. If the botnet is blasting the business with 150Mbits/s of spoofed traffic and the connection maxes out at 100Mbits/s, then genuine traffic will be submerged. Performance will be so compromised that the connection will be unusable.
But even if the typical bandwidth needs are high, it’s possible the total amount of packets per second (pps) will exceed the ability of the router to process. Equipment specified to handle 20,000pps may find itself trying to deal with 200,000pps. The reality is few companies have equipment able to handle a huge unexpected spike, which means genuine traffic won’t be able to get through, Vector says.
So, what can Kiwi companies do to survive such an attack?
To minimise long-term damage, businesses need to prepare for the possibility of a DDoS and have robust – and tested – plans in place to mitigate the effects, according to Vector. Even if a business is not deemed to be ‘high profile’ they could still be a target as the motivations of those who launch DDoS attacks are increasingly unpredictable.
Vector has laid out three steps to help stop DDoS attacks:
You’re under attack - what next?
The aim is stop the attack as soon as possible, and the best place to do that is at your upstream ISP provider so the traffic is eliminated before it hits the business connection, Vector says.
If the provider hasn’t already detected the DDoS (which they may do if it is severe enough), it is important for the business to contact their network operations centre (NOC) and give them the news. Then they can either scrub the traffic or block it, Vector says.
Scrubbing involves analysing the traffic and if there’s anything suspicious that looks like it’s come from a DDoS attack, it is quietly dropped (scrubbed) while legitimate traffic is untouched. Not all providers have the ability to scrub traffic and those that do may charge extra for the service so check the options and potential costs as part of planning, Vector says.
If the connection to the ISP is BGP-based (Border Gateway Protocol), and the ISP supports BGP communities that automatically null-route IPs with a specific community tag (Vector says businesses should ask their provider if they do this), then the business can start advertising a /32 route with that tag to the upstream and they will automatically block it. This can be rapid, as it is semi-automated, and won’t require a human response from your provider’s NOC, according to Vector.
If a provider can’t scrub the traffic then blocking is the next best option. It will result in the business network being completely blocked, which is obviously going to result in reduced internet availability, but it will end the DDoS attack, Vector says.
If the scale of the attack is big enough the provider may block traffic without a request from the business; if the packet-per-second level is high enough to saturate their links they’ll act to protect themselves and other customers.
Once you’ve taken these steps, the business has to simply wait it out. A DDoS attack isn’t an infinitely sustainable tactic, so the longer it runs, the more responders learn about the botnets being used to launch it and the more focused the response will become.
Although major attacks lasting days have occurred, the majority of DDoS attacks are of a smaller scale and last from a few minutes to several hours, Vector says.
“Surviving a DDoS attack is all about getting your planning right. You can never eliminate the chance of an attack happening but, if you are targeted, the right mitigation plan - created and tested with your ISP - means you’ll be well-positioned to ride out the storm,” Vector says.