By design, BIND is intended to limit the number of TCP clients that can be connected at any given time. The number of allowed connections is a tunable parameter which, if unset, defaults to a conservative value for most servers. Unfortunately, the code which was intended to limit the number of simultaneous connections contained an error which could be exploited to grow the number of simultaneous connections beyond this limit. Versions affected: BIND 9.9.0 -> 9.10.8-P1, 9.11.0 -> 9.11.6, 9.12.0 -> 9.12.4, 9.14.0. BIND 9 Supported Preview Edition versions 9.9.3-S1 -> 9.11.5-S3, and 9.11.5-S5. Versions 9.13.0 -> 9.13.7 of the 9.13 development branch are also affected. Versions prior to BIND 9.9.0 have not been evaluated for vulnerability to CVE-2018-5743.

A query with a specific set of characteristics could cause a server using DNS64 to encounter an assertion failure and terminate. An attacker could deliberately construct a query, enabling denial-of-service against a server if it was configured to use the DNS64 feature and other preconditions were met.

Servers are at risk if they are configured to use DNS64 and if the option "break-dnssec yes;" is in use.

External References:

https://kb.isc.org/article/AA-01465

Mitigation:

Servers which have configurations which require DNS64 and "break-dnssec yes;" should upgrade. Servers which are not using these features in conjunction are not at risk from this defect.

BIND was improperly sequencing cleanup operations on upstream recursion fetch contexts, leading in some cases to a use-after-free error that can trigger an assertion failure and crash in named. Affects BIND 9.0.0 to 9.8.x, 9.9.0 to 9.9.11, 9.10.0 to 9.10.6, 9.11.0 to 9.11.2, 9.9.3-S1 to 9.9.11-S1, 9.10.5-S1 to 9.10.6-S1, 9.12.0a1 to 9.12.0rc1.

A flaw was found in Bind. Controls for zone transfers may not be properly applied to Dynamically Loadable Zones (DLZs) if the zones are writable. A client exercising this defect can request and receive a zone transfers of a DLZ even when not permitted to do so by the allow-transfer ACL.

The BIND installer on Windows uses an unquoted service path which can enable a local user to achieve privilege escalation if the host file system permissions allow this. Affects BIND 9.2.6-P2->9.2.9, 9.3.2-P1->9.3.6, 9.4.0->9.8.8, 9.9.0->9.9.10, 9.10.0->9.10.5, 9.11.0->9.11.1, 9.9.3-S1->9.9.10-S1, 9.10.5-S1.

An attacker who is able to send and receive messages to an authoritative DNS server and who has knowledge of a valid TSIG key name for the zone and service being targeted may be able to manipulate BIND into accepting an unauthorized dynamic update. A server that relies solely on TSIG or SIG(0) keys with no other address-based ACL protection could be vulnerable to malicious zone content manipulation using this technique.

Workarounds:

The effects of this vulnerability can be mitigated by using Access Control Lists (ACLs) that require both address range validation and use of TSIG authentication in parallel. For information on how to configure this type of compound authentication control, please see:

https://kb.isc.org/article/AA-00723/0/Using-Access-Control-Lists-ACLs-with-both-addresses-and-keys.html.

Administrators who have made use of named.conf option "update-policy local;" should refer to the Administrator Reference Manual (ARM) for details of the automatic update policy that will be established and to assess whether or not this conveys any additional risk to their server. (Note that this option is not enabled by default).

Upstream patch:

https://source.isc.org/cgi-bin/gitweb.cgi?p=bind9.git;a=commitdiff;h=581c1526ab

ISC BIND 9.8.x through 9.8.4-P1 and 9.9.x through 9.9.2-P1, in certain configurations involving DNS64 with a Response Policy Zone that lacks an AAAA rewrite rule, allows remote attackers to cause a denial of service (assertion failure and named daemon exit) via a query for an AAAA record.

ISC BIND 9.8.x before 9.8.4-P1 and 9.9.x before 9.9.2-P1, when DNS64 is enabled, allows remote attackers to cause a denial of service (assertion failure and daemon exit) via a crafted query.

ISC BIND 9.x before 9.7.6-P4, 9.8.x before 9.8.3-P4, 9.9.x before 9.9.1-P4, and 9.4-ESV and 9.6-ESV before 9.6-ESV-R7-P4 allows remote attackers to cause a denial of service (named daemon hang) via unspecified combinations of resource records.

ISC BIND 9.x before 9.7.6-P3, 9.8.x before 9.8.3-P3, 9.9.x before 9.9.1-P3, and 9.4-ESV and 9.6-ESV before 9.6-ESV-R7-P3 allows remote attackers to cause a denial of service (assertion failure and named daemon exit) via a query for a long resource record.

Race condition in the nsclient structure management in ISC BIND 9.9.x before 9.9.1-P2 allows remote attackers to cause a denial of service (memory consumption or process exit) via a large volume of TCP queries.

ISC BIND 9.4.x, 9.5.x, 9.6.x, and 9.7.x before 9.7.6-P2; 9.8.x before 9.8.3-P2; 9.9.x before 9.9.1-P2; and 9.6-ESV before 9.6-ESV-R7-P2, when DNSSEC validation is enabled, does not properly initialize the failing-query cache, which allows remote attackers to cause a denial of service (assertion failure and daemon exit) by sending many queries.

Cache Poisoning issue exists in DNS Response Rate Limiting.

The RFC 5011 implementation in rdata.c in ISC BIND 9.7.x and 9.8.x before 9.8.5-P2, 9.8.6b1, 9.9.x before 9.9.3-P2, and 9.9.4b1, and DNSco BIND 9.9.3-S1 before 9.9.3-S1-P1 and 9.9.4-S1b1, allows remote attackers to cause a denial of service (assertion failure and named daemon exit) via a query with a malformed RDATA section that is not properly handled during construction of a log message, as exploited in the wild in July 2013.

libdns in ISC BIND 9.7.x and 9.8.x before 9.8.4-P2, 9.8.5 before 9.8.5b2, 9.9.x before 9.9.2-P2, and 9.9.3 before 9.9.3b2 on UNIX platforms allows remote attackers to cause a denial of service (memory consumption) via a crafted regular expression, as demonstrated by a memory-exhaustion attack against a machine running a named process.

ISC BIND 9.0.x through 9.8.x, 9.9.0 through 9.9.6, and 9.10.0 through 9.10.1 does not limit delegation chaining, which allows remote attackers to cause a denial of service (memory consumption and named crash) via a large or infinite number of referrals.

Race condition in resolver.c in named in ISC BIND 9.9.8 before 9.9.8-P2 and 9.10.3 before 9.10.3-P2 allows remote attackers to cause a denial of service (INSIST assertion failure and daemon exit) via unspecified vectors.

db.c in named in ISC BIND 9.x before 9.9.8-P2 and 9.10.x before 9.10.3-P2 allows remote attackers to cause a denial of service (REQUIRE assertion failure and daemon exit) via a malformed class attribute.

name.c in named in ISC BIND 9.7.x through 9.9.x before 9.9.7-P1 and 9.10.x before 9.10.2-P2, when configured as a recursive resolver with DNSSEC validation, allows remote attackers to cause a denial of service (REQUIRE assertion failure and daemon exit) by constructing crafted zone data and then making a query for a name in that zone.

It was reported that a problem with trust anchor management can cause named to crash, affecting BIND versions 9.7.0+.

ISC developers believe that it will be very difficult for this to be triggered in most cases, requiring DNSSEC validation amongst other factors.

ISC will not be producing patches specifically for BIND 9.8 or BIND 9.6-ESV, both of which are beyond their End of Life (EOL) and are no longer supported by ISC.

Patches that correct this issue for ISC BIND 9.10.1 and ISC BIND 9.9.6 are attached to this Bugzilla.

A packet with a malformed options section can be used to deliberately trigger an assertion failure affecting versions of BIND which do not contain change #3548.

A server vulnerable to this defect can be forced to exit with an assertion failure if it receives a malformed packet. Authoritative and recursive servers are both vulnerable.

https://source.isc.org/cgi-bin/gitweb.cgi?p=bind9.git;a=blob;f=CHANGES has more information on change #3548. The commit corresponding to this change is https://source.isc.org/cgi-bin/gitweb.cgi?p=bind9.git;a=commitdiff;h=4adf97c32fcca7d00e5756607fd045f2aab9c3d4.