For wireless and wireline communications service providers (CSPs), subscribers are accessing content more than ever before, forcing network operators to support higher and higher levels of traffic. Top on the subscriber’s mind is performance. As we are all encountering (especially more and more), use cases such as video streaming and online meetings are susceptible to network latency.
CSPs are moving services closer to customers, often right to the edge of the network to maintain low latency, high quality, and excellent customer experience. But for this strategy to succeed, they also need to move their Domain Name System (DNS) servers to the Network Edge to ensure low DNS latency. DNS is a foundational Internet protocol that converts human-readable names to IP addresses, changes IP addresses back to names, and provides easy-to-remember names for many Internet-based services.
Everything online begins with a DNS request. Without fast DNS responses, the subscriber experience suffers. And ultimately, performance won’t be good enough to keep subscribers coming back. While moving DNS to the edge of the network seems like a great idea, it adds a lot of complexity for the network operations teams who have to support the DNS infrastructure. Supporting tens or even hundreds of remote servers is challenging, especially given that generic DNS servers need very frequent updates to software, hardware, and operating systems, especially in the mobile world.
Then there’s 5G and Multi-Access Edge Computing (MEC). The pervasive connectivity of 5G will increase reliance on edge computing, bringing cloud resources such as compute, storage, and networking—closer to applications, devices, and users. In fact, 5G implementations will require greater use of small cell stations at the very edge of the network so that data will not need to travel long distances to a cloud or data center. To ensure unhindered traffic flow at the edge, DNS services must also be positioned at the edge. The result? The higher number of edge-based applications and 5G radio deployments will increase the number of DNS servers that network operators will need to manage.
DNS Query and Response Logging
Understanding what is going on within the network is essential to service providers, and DNS plays a critical role in gathering that intelligence. Clearly, from both a performance and security standpoint, knowing what subscriber devices are looking up in DNS provides valuable information about healthy versus abnormal operations. DNS query and response logging has been one of the significant ways for a DNS system to produce raw data on what queries are being made. At the same time, reporting is the organization and transformation of that raw data into humanly readable formats.
As a DNS server sends queries to other DNS servers, it can keep a copy of these queries for later analysis or aid in troubleshooting DNS-related issues. Also, the information can be processed later to generate reports with information such as “What was the most queried domain name in the last hour?”. System log or syslog is typically the mechanism to record this query information, a process known as query logging. Similarly, as a DNS server receives answers, besides sending the responses to the client, that server could also keep a copy of the answer for later analysis or to aid in troubleshooting. It can use the same mechanism as query logging by storing this information in logs, which is known as response logging.
But here’s the problem. While query and response logging provide a valuable source of data for both security and operational analysis, both processes can significantly add more I/O (Input/Output) load to the DNS servers and are typically disabled by default in operational environments. With query logging, every outbound query sent by the DNS server translates to (eventually) another write-operation to the disk. This load is even more burdensome when response logging is enabled since the number of responses could be very large. The speed of disk storage is a limiting factor, with the additional load on I/O resource-limited hosts, query processing must wait until the log data has been written to disk.
That can be painful, considering that network performance is everything to the subscriber.
High-Speed Query Logging with dnstap
Infoblox Network Identity Operating System (NIOS) is the operating system that powers Infoblox core network services, ensuring the non-stop operation of network infrastructure. In our latest version of NIOS, we are adding high-speed query logging through dnstap. dnstap (the specifications can be found through the link) is a relatively new and open standard for how to monitor and look deeper into what DNS servers are doing. It’s meant for large and active DNS server implementations characteristic of network operators. Designed to be fast and lightweight, it provides a quick, flexible method for capturing and logging DNS traffic without significant performance tradeoff. Unlike I/O-intensive query and response logging, dnstap operates asynchronously, allowing high-speed query logging with minimal performance loss.
Available on the Infoblox DNS Cache Acceleration solution, dnstap uses Google’s protocol buffer implementation to package log query and response information directly from the DNS server within a flexible binary structured log format, and then stream this data to a receiving server. From there, service providers can leverage big data tools to analyze this data (and even combine this data with other sources) to build a complete picture of subscriber usage patterns based on network data.
Digging Deeper: dnstap Use Cases
The more information that service providers have about their networks, the better they can manage and protect not just the network, but the business. Many service providers and enterprises already utilize Infoblox Reporting and Analytics to takes this data and turn it into Actionable Network Intelligence—enabling them to create custom dashboards and reports to troubleshoot network and application problems, uncover security threats, and address compliance requirements.
Service providers are unique, since the network IS their business, so they sometimes require more granular data on subscriber usage patterns to facilitate network and product planning. Subscriber data can be likened to gold for service providers, and dnstap can provide a veritable firehose of subscriber behavioral data. Here are a few use cases that come to mind:
- Uncover subscriber impacts. Here’s a great example from recent events: how has online subscriber behaviors changed before and after social distancing restrictions. Network data can highlight how subscriber usage patterns have changed, and what actions can be taken to ensure optimal network performance. Another example: what is the impact to subscribers based on a new strain of malware? Again, service providers can determine effects based on malware DNS queries.
- Protect the brand. Service providers are continually seeking ways to combat customer churn in the face of an ultra-competitive environment. With DNS network data, it is now possible to model subscriber behaviors and identify those at risk to churn. Are subscribers viewing competitor websites? Are they leaving online reviews of their experiences on various properties? Network data with usage profiles is a critical input into developing predictive churn models, and a great indicator of the subscriber usage experience.
- Generate new revenue streams. Depending on location, recent changes to regulatory environments may have opened the door for service providers to increase average revenue per user (ARPU) by monetizing the data flowing through their networks. Service providers can identify trends in subscriber network data and leverage this information to develop new product offerings and enhancements and even expand their offerings into newer areas such as IoT. When combined with Infoblox Subscriber Services, service providers can gain leverage data from the individual subscriber and device level to help develop targeted new services to benefit subscribers and increase revenues.
Infoblox Provides Visibility and Performance
Performance matters more than ever. Service provider DNS caching infrastructure needs to be robust. When subscribers experience slow page loads, buffered video, or suspicious app activity, the overall experience is affected—and loyalty to their current service provider may be questioned. Infoblox has provided carrier-grade security, high availability, and ultra-low latency that helps create a better first-connection impression for subscribers. With dnstap, we’re giving service providers even greater visibility and insight into their DNS servers while retaining the optimal ultra-low latency they expect.
NIOS 8.5.1 information:
Product page: https://www.infoblox.com/products/nios8/
Solution Note: https://www.infoblox.com/resources/solution-notes/nios-8
Service Provider page: https://www.infoblox.com/solutions/service-providers/
DNS Caching page: https://www.infoblox.com/solutions/service-providers/secure-dns-caching/