Login
The Domain Name System (DNS) root zone will soon be getting a new record type, called ZONEMD, to further ensure the security, stability, and resiliency of the global DNS in the face of emerging new approaches to DNS operation. While this change will be unnoticeable for the vast majority of DNS operators (such as registrars, internet service providers, and organizations), it provides a valuable additional layer of cryptographic security to ensure the reliability of root zone data.
In this blog, we’ll discuss these new proposals, as well as ZONEMD. We’ll share deployment plans, how they may affect certain users, and what DNS operators need to be aware of beforehand to ensure little-to-no disruptions.
The DNS root zone is the starting point for most domain name lookups on the internet. The root zone contains delegations to nearly 1,500 top-level domains, such as .com, .net, .org, and many others. Since its inception in 1984, various organizations known collectively as the Root Server Operators have provided the service for what we now call the Root Server System (RSS). In this system, a myriad of servers respond to approximately 80 billion root zone queries each day.
While the RSS continues to perform this function with a high degree of dependability, there are recent proposals to use the root zone in a slightly different way. These proposals create some efficiencies for DNS operators, but they also introduce new challenges.
In 2020, the Internet Engineering Task Force (IETF) published RFC 8806, titled “Running a Root Server Local to a Resolver.” Along the same lines, in 2021 the Internet Corporation for Assigned Names and Numbers (ICANN) Office of the Chief Technology Officer published OCTO-027, titled “Hyperlocal Root Zone Technical Analysis.” Both proposals share the idea that recursive name servers can receive and load the entire root zone locally and respond to root zone queries directly.
But in a scenario where the entire root zone is made available to millions of recursive name servers, a new question arises: how can consumers of zone data verify that zone content has not been modified before reaching their systems?
One might imagine that DNS Security Extensions (DNSSEC) could help. However, while the root zone is indeed signed with DNSSEC, most of the records in the zone are considered non-authoritative (i.e., all the NS and glue records) and therefore do not have signatures. What about something like a Pretty Good Privacy (PGP) signature on the root zone file? That comes with its own challenge: in PGP, the detached signature is easily separated from the data. For example, there is no way to include a PGP signature over DNS zone transfer, and there is no easy way to know which version of the zone goes with the signature.
A solution to this problem comes from RFC 8976. Led by Verisign and titled “Message Digest for DNS Zones” (known colloquially as ZONEMD), this protocol calls for a cryptographic digest of the zone data to be embedded into the zone itself. This ZONEMD record can then be signed and verified by consumers of the zone data. Here’s how it works:
Each time a zone is updated, the publisher calculates the ZONEMD record by sorting and canonicalizing all the records in the zone and providing them as input to a message digest function. Sorting and canonicalization are the same as for DNSSEC. In fact, the ZONEMD calculation can be performed at the same time the zone is signed. Digest calculation necessarily excludes the ZONEMD record itself, so the final step is to update the ZONEMD record and its signatures.
A recipient of a zone that includes a ZONEMD record repeats the same calculation and compares its calculated digest value with the published digest. If the zone is signed, then the recipient can also validate the correctness of the published digest. In this way, recipients can verify the authenticity of zone data before using it.
A number of open-source DNS software products now, or soon will, include support for ZONEMD verification. These include Unbound (version 1.13.2), NSD (version 4.3.4), Knot DNS (version 3.1.0), PowerDNS Recursor (version 4.7.0) and BIND (version 9.19).
Verisign, ICANN, and the Root Server Operators are taking steps to ensure that the addition of the ZONEMD record in no way impacts the ability of the root server system to receive zone updates and to respond to queries. As a result, most internet users are not affected by this change.
Anyone using RFC 8806, or a similar technique to load root zone data into their local resolver, is unlikely to be affected as well. Software products that implement those features should be able to fully process a zone that includes the new record type, especially for reasons described below. Once the record has been added, users can take advantage of ZONEMD verification to ensure root zone data is authentic.
Users most likely to be affected are those that receive root zone data from the internic.net servers (or some other source) and use custom software to parse the zone file. Depending on how such custom software is designed, there is a possibility that it will treat the new ZONEMD record as unexpected and lead to an error condition. Key objectives of this blog post are to raise awareness of this change, provide ample time to address software issues, and minimize the likelihood of disruptions for such users.
In 2020, Verisign asked the Root Zone Evolution Review Committee (RZERC) to consider a proposal for adding data protections to the root zone using ZONEMD. In 2021, the RZERC published its recommendations in RZERC003. One of those recommendations was for Verisign and ICANN to develop a deployment plan and make the community aware of the plan’s details. That plan is summarized in the remainder of this blog post.
One attribute of a ZONEMD record is the choice of a hash algorithm used to create the digest. RFC 8976 defines two standard hash algorithms – SHA-384 and SHA-512 – and a range of “private-use” algorithms.
Initially, the root zone’s ZONEMD record will have a private-use hash algorithm. This allows us to first include the record in the zone without anyone worrying about the validity of the digest values. Since the hash algorithm is from the private-use range, a consumer of the zone data will not know how to calculate the digest value. A similar technique, known as the “Deliberately Unvalidatable Root Zone,” was utilized when DNSSEC was added to the root zone in 2010.
After a period of more than two months, the ZONEMD record will transition to a standard hash algorithm.
SHA-384 has been selected for the initial implementation for compatibility reasons.
The developers of BIND implemented the ZONEMD protocol based on an early Internet-Draft, some time before it was published as an RFC. Unfortunately, the initial BIND implementation only accepts ZONEMD records with a digest length of 48 bytes (i.e., the SHA-384 length). Since the versions of BIND with this behavior are in widespread use today, use of the SHA-512 hash algorithm would likely lead to problems for many BIND installations, possibly including some Root Server Operators.
Distribution of the zone between the Root Zone Maintainer and Root Server Operators primarily takes place via the DNS zone transfer protocol. In this protocol, zone data is transmitted in “wire format.”
The root zone is also stored and served as a file on the internic.net FTP and web servers. Here, the zone data is in “presentation format.” The ZONEMD record will appear in these files using its native presentation format. For example:
FreshRSS 
. 86400 IN ZONEMD 2021101902 1 1 ( 7d016e7badfd8b9edbfb515deebe7a866bf972104fa06fec
e85402cc4ce9b69bd0cbd652cec4956a0f206998bfb34483 )Some users of zone data received from the FTP and web servers might currently be using software that does not recognize the ZONEMD presentation format. These users might experience some problems when the ZONEMD record first appears. We did consider using a generic record format; however, in consultation with ICANN, we believe that the native format is a better long-term solution.
Currently, we are targeting the initial deployment of ZONEMD in the root zone for September 13, 2023. As previously stated, the ZONEMD record will be published first with a private-use hash algorithm number. We are targeting December 6, 2023, as the date to begin using the SHA-384 hash algorithm, at which point the root zone ZONEMD record will become verifiable.
Deploying ZONEMD in the root zone helps to increase the security, stability, and resiliency of the DNS. Soon, recursive name servers that choose to serve root zone data locally will have stronger assurances as to the zone’s validity.
If you’re interested in following the ZONEMD deployment progress, please look for our announcements on the DNS Operations mailing list.
The post Adding ZONEMD Protections to the Root Zone appeared first on Verisign Blog.
Every few months, an important ceremony takes place. It’s not splashed all over the news, and it’s not attended by global dignitaries. It goes unnoticed by many, but its effects are felt across the globe. This ceremony helps make the internet more secure for billions of people.
This unique ceremony began in 2010 when Verisign, the Internet Corporation for Assigned Names and Numbers (ICANN), and the U.S. Department of Commerce’s National Telecommunications and Information Administration collaborated – with input from the global internet community – to deploy a technology called Domain Name System Security Extensions (DNSSEC) to the Domain Name System (DNS) root zone in a special ceremony. This wasn’t a one-off occurrence in the history of the DNS, though. Instead, these organizations developed a set of processes, procedures, and schedules that would be repeated for years to come. Today, these recurring ceremonies help ensure that the root zone is properly signed, and as a result, the DNS remains secure, stable, and resilient.
In this blog, we take the opportunity to explain these ceremonies in greater detail and describe the critical role that Verisign is honored to perform.
DNSSEC is a series of technical specifications that allow operators to build greater security into the DNS. Because the DNS was not initially designed as a secure system, DNSSEC represented an essential leap forward in securing DNS communications. Deploying DNSSEC allows operators to better protect their users, and it helps to prevent common threats such as “man-in-the-middle” attacks. DNSSEC works by using public key cryptography, which allows zone operators to cryptographically sign their zones. This allows anyone communicating with and validating a signed zone to know that their exchanges are genuine.
The root zone, like most signed zones, uses separate keys for zone signing and for key signing. The Key Signing Key (KSK) is separate from the Zone Signing Key (ZSK). However, unlike most zones, the root zone’s KSK and ZSK are operated by different organizations; ICANN serves as the KSK operator and Verisign as the ZSK operator. These separate roles for DNSSEC align naturally with ICANN as the Root Zone Manager and Verisign as the Root Zone Maintainer.
In practice, the KSK/ZSK split means that the KSK only signs the DNSSEC keys, and the ZSK signs all the other records in the zone. Signing with the KSK happens infrequently – only when the keys change. However, signing with the ZSK happens much more frequently – whenever any of the zone’s other data changes.
Something to keep in mind before we go further: remember that DNSSEC utilizes public key cryptography, in which keys have both a private and public component. The private component is used to generate signatures and must be guarded closely. The public component is used to verify signatures and can be shared openly. Good cryptographic hygiene says that these keys should be changed (or “rolled”) periodically.
In DNSSEC, changing a KSK is generally difficult, whereas changing a ZSK is relatively easy. This is especially true for the root zone where a KSK rollover requires all validating recursive name servers to update their copy of the trust anchor. Whereas the first and only KSK rollover to date happened after a period of eight years, ZSK rollovers take place every three months. Not coincidentally, this is also how often root zone key signing ceremonies take place.
The notion of holding a “ceremony” for such an esoteric technical function may seem strange, but this ceremony is very different from what most people are used to. Our common understanding of the word “ceremony” brings to mind an event with speeches and formal attire. But in this case, the meaning refers simply to the formality and ritual aspects of the event.
There are two main reasons for holding key signing ceremonies. One is to bring participants together so that everyone may transparently witness the process. Ceremony participants include ICANN staff, Verisign staff, Trusted Community Representatives (TCRs), and external auditors, plus guests on occasion.
The other important reason, of course, is to generate DNSSEC signatures. Occasionally other activities take place as well, such as generating new keys, retiring equipment, and changing TCRs. In this post, we’ll focus only on the signature generation procedures.
A month or two before each ceremony, Verisign generates a file called the Key Signing Request (KSR). This is an XML document which includes the set of public key records (both KSK and ZSK) to be signed and then used during the next calendar quarter. The KSR is securely transmitted from Verisign to the Internet Assigned Numbers Authority (IANA), which is a function of ICANN that performs root zone management. IANA securely stores the KSR until it is needed for the upcoming key signing ceremony.
Each quarter is divided into nine 10-day “slots” (for some quarters, the last slot is extended by a day or two) and the XML file contains nine key “bundles” to be signed. Each bundle, or slot, has a signature inception and expiration timestamp, such that they overlap by at least five days. The first and last slots in each quarter are used to perform ZSK rollovers. During these slots we publish two ZSKs and one KSK in the root zone.
The root zone KSK private component is held inside secure Hardware Security Modules (HSMs). These HSMs are stored inside locked safes, which in turn are kept inside locked rooms. At a key signing ceremony, the HSMs are taken out of their safes and activated for use. This all occurs according to a pre-defined script with many detailed steps, as shown in the figure below.
Also stored inside the safe is a laptop computer, its operating system on non-writable media (i.e., DVD), and a set of credentials for the TCRs, stored on smart cards and locked inside individual safe deposit boxes. Once all the necessary items are removed from the safes, the equipment can be turned on and activated.
The laptop computer is booted from its operating system DVD and the HSM is connected via Ethernet for data transfer and serial port for console logging. The TCR credentials are used to activate the HSM. Once activated, a USB thumb drive containing the KSR file is connected to the laptop and the signing program is started.
The signing program reads the KSR, validates it, and then displays information about the keys about to be signed. This includes the signature inception and expiration timestamps, and the ZSK key tag values.
Validate and Process KSR /media/KSR/KSK46/ksr-root-2022-q4-0.xml...
# Inception Expiration ZSK Tags KSK Tag(CKA_LABEL)
1 2022-10-01T00:00:00 2022-10-22T00:00:00 18733,20826
2 2022-10-11T00:00:00 2022-11-01T00:00:00 18733
3 2022-10-21T00:00:00 2022-11-11T00:00:00 18733
4 2022-10-31T00:00:00 2022-11-21T00:00:00 18733
5 2022-11-10T00:00:00 2022-12-01T00:00:00 18733
6 2022-11-20T00:00:00 2022-12-11T00:00:00 18733
7 2022-11-30T00:00:00 2022-12-21T00:00:00 18733
8 2022-12-10T00:00:00 2022-12-31T00:00:00 18733
9 2022-12-20T00:00:00 2023-01-10T00:00:00 00951,18733
...PASSED.
It also displays an SHA256 hash of the KSR file and a corresponding “PGP (Pretty Good Privacy) Word List.” The PGP Word List is a convenient and efficient way of verbally expressing hexadecimal values:
SHA256 hash of KSR:
ADCE9749F3DE4057AB680F2719B24A32B077DACA0F213AD2FB8223D5E8E7CDEC
>> ringbolt sardonic preshrunk dinosaur upset telephone crackdown Eskimo rhythm gravity artist celebrate bedlamp pioneer dogsled component ruffled inception surmount revenue artist Camelot cleanup sensation watchword Istanbul blowtorch specialist trauma truncated spindle unicorn <<
At this point, a Verisign representative comes forward to verify the KSR. The following actions then take place:
The signing program outputs a new XML document, called the Signed Key Response (SKR). This document contains signatures over the DNSKEY resource record sets in each of the nine slots. The SKR is saved to a USB thumb drive and given to a member of the Root Zone KSK Operations Security team. Usually sometime the next day, IANA securely transmits the SKR back to Verisign. Following several automatic and manual verification steps, the signature data is imported into Verisign’s root zone management system for use at the appropriate times in the next calendar quarter.
Keeping the internet’s DNS secure, stable, and resilient is a crucial aspect of Verisign’s role as the Root Zone Maintainer. We are honored to participate in the key signing ceremonies with ICANN and the TCRs and do our part to help the DNS operate as it should.
For more information on root key signing ceremonies, visit the IANA website. Visitors can watch video recordings of previous ceremonies and even sign up to witness the next ceremony live. It’s a great resource, and a unique opportunity to take part in a process that helps keep the internet safe for all.
The post Verisign’s Role in Securing the DNS Through Key Signing Ceremonies appeared first on Verisign Blog.
This blog was also published by APNIC.
With so much traffic on the global internet day after day, it’s not always easy to spot the occasional irregularity. After all, there are numerous layers of complexity that go into the serving of webpages, with multiple companies, agencies and organizations each playing a role.
That’s why when something does catch our attention, it’s important that the various entities work together to explore the cause and, more importantly, try to identify whether it’s a malicious actor at work, a glitch in the process or maybe even something entirely intentional.
That’s what occurred last year when Internet Corporation for Assigned Names and Numbers staff and contractors were analyzing names in Domain Name System queries seen at the ICANN Managed Root Server, and the analysis program ran out of memory for one of their data files. After some investigating, they found the cause to be a very large number of mysterious queries for unique names such as f863zvv1xy2qf.surgery, bp639i-3nirf.hiphop, qo35jjk419gfm.net and yyif0aijr21gn.com.
While these were queries for names in existing top-level domains, the first label consisted of 12 or 13 random-looking characters. After ICANN shared their discovery with the other root server operators, Verisign took a closer look to help understand the situation.
One of the first things we noticed was that all of these mysterious queries were of type NS and came from one autonomous system network, AS 15169, assigned to Google LLC. Additionally, we confirmed that it was occurring consistently for numerous TLDs. (See Fig. 1)
Although this phenomenon was newly uncovered, analysis of historical data showed these traffic patterns actually began in late 2019. (See Fig. 2)
Perhaps the most interesting discovery, however, was that these specific query names were not also seen at the .com and .net name servers operated by Verisign. The data in Figure 3 shows the fraction of queried names that appear at A-root and J-root and also appear on the .com and .net name servers. For second-level labels of 12 and 13 characters, this fraction is essentially zero. The graphs also show that there appears to be queries for names with second-level label lengths of 10 and 11 characters, which are also absent from the TLD data.
The final mysterious aspect to this traffic is that it deviated from our normal expectation of caching. Remember that these are queries to a root name server, which returns a referral to the delegated name servers for a TLD. For example, when a root name server receives a query for yyif0aijr21gn.com, the response is a list of the name servers that are authoritative for the .com zone. The records in this response have a time to live of two days, meaning that the recursive name server can cache and reuse this data for that amount of time.
However, in this traffic we see queries for .com domain names from AS 15169 at the rate of about 30 million per day. (See Fig. 4) It is well known that Google Public DNS has thousands of backend servers and limits TTLs to a maximum of six hours. Assuming 4,000 backend servers each cached a .com referral for six hours, we might expect about 16,000 queries over a 24-hour period. The observed count is about 2,000 times higher by this back-of-the-envelope calculation.
From our initial analysis, it was unclear if these queries represented legitimate end-user activity, though we were confident that source IP address spoofing was not involved. However, since the query names shared some similarities to those used by botnets, we could not rule out malicious activity.
These findings were presented last year at the DNS-OARC 35a virtual meeting. In the conference chat room after the talk, the missing piece of this puzzle was mentioned by a conference participant. There is a Google webpage describing its public DNS service that talks about prepending nonce (i.e., random) labels for cache misses to increase entropy. In what came to be known as “the Kaminsky Attack,” an attacker can cause a recursive name server to emit queries for names chosen by the attacker. Prepending a nonce label adds unpredictability to the queries, making it very difficult to spoof a response. Note, however, that nonce prepending only works for queries where the reply is a referral.
In addition, Google DNS has implemented a form of query name minimization (see RFC 7816 and RFC 9156). As such, if a user requests the IP address of www.example.com and Google DNS decides this warrants a query to a root name server, it takes the name, strips all labels except for the TLD and then prepends a nonce string, resulting in something like u5vmt7xanb6rf.com. A root server’s response to that query is identical to one using the original query name.
Now, we are able to explain nearly all of the mysterious aspects of this query traffic from Google. We see random second-level labels because of the nonce strings that are designed to prevent spoofing. The 12- and 13-character-long labels are most likely the result of converting a 64-bit random value into an unpadded ASCII label with encoding similar to Base32. We don’t observe the same queries at TLD name servers because of both the nonce prepending and query name minimization. The query type is always NS because of query name minimization.
With that said, there’s still one aspect that eludes explanation: the high query rate (2000x for .com) and apparent lack of caching. And so, this aspect of the mystery continues.
Even though we haven’t fully closed the books on this case, one thing is certain: without the community’s teamwork to put the pieces of the puzzle together, explanations for this strange traffic may have remained unknown today. The case of the mysterious DNS root query traffic is a perfect example of the collaboration that’s required to navigate today’s ever-changing cyber environment. We’re grateful and humbled to be part of such a dedicated community that is intent on ensuring the security, stability and resiliency of the internet, and we look forward to more productive teamwork in the future.
The post More Mysterious DNS Root Query Traffic from a Large Cloud/DNS Operator appeared first on Verisign Blog.
