Note that the Rate Limiting option has been renamed.
Security Fixes
Previously an error in bounds checking on the private type
'keydata' could be used to deny service through a deliberately
triggerable REQUIRE failure (CVE-2013-4854). [RT #34238]
Prevents exploitation of a runtime_check which can crash named
when satisfying a recursive query for particular malformed zones.
(CVE-2013-3919) [RT #33690]
New Features
Added Response Rate Limiting (RRL) functionality to reduce the
effectiveness of DNS as an amplifier for reflected denial-of-service
attacks by rate-limiting substantially-identical responses. [RT
#28130]
Feature Changes
rndc status now also shows the build-id. [RT #20422]
Improved OPT pseudo-record processing to make it easier to support
new EDNS options. [RT #34414]
"configure" now finishes by printing a summary of optional BIND
features and whether they are active or inactive. ("configure
--enable-full-report" increases the verbosity of the summary.)
[RT #31777]
Addressed compatibility issues with newer versions of Microsoft
Visual Studio. [RT #33916]
Improved the 'rndc' man page. [RT #33506]
'named -g' now no longer works with an invalid logging configuration.
[RT #33473]
The default (and minimum) value for tcp-listen-queue is now 10
instead of 3. This is a subtle control setting (not applicable
to all OS environments). When there is a high rate of inbound
TCP connections, it controls how many connections can be queued
before they are accepted by named. Once this limit is exceeded,
new TCP connections will be rejected. Note however that a value
of 10 does not imply a strict limit of 10 queued TCP connections
- the impact of changing this configuration setting will be
OS-dependent. Larger values for tcp-listen queue will permit
more pending tcp connections, which may be needed where there
is a high rate of TCP-based traffic (for example in a dynamic
environment where there are frequent zone updates and transfers).
For most production servers the new default value of 10 should
be adequate. [RT #33029]
Added support for OpenSSL versions 0.9.8y, 1.0.0k, and 1.0.1e
with PKCS#11. [RT #33463]
Added logging messages on slave servers when they forward DDNS
updates to a master. [RT #33240]
Changed the logging category for RRL events from 'queries' to
'query-errors'. [RT #33540]
"--with-libiconv=${LOCALBASE}" at systems pre OSVERSION 100043 and "" (null)
otherwise;
. convert all ports which has CONFIGURE_ARGS=--with-libiconv=${LOCALBASE}.
Approved by: portmgr (bapt, implicit)
https://kb.isc.org/article/AA-01015/0
9.9.3-p1 -> 9.9.3-P2
9.8.5-p1 -> 9.8.5-P2
9.6.x is not affected, neither is 10.x.
Security: CVE-2013-4854 Remote DOS
Security Fixes
Prevents exploitation of a runtime_check which can crash named
when satisfying a recursive query for particular malformed zones.
(CVE-2013-3919) [RT #33690]
Now supports NAPTR regular expression validation on all platforms,
and avoids memory exhaustion compiling pathological regular
expressions. (CVE-2013-2266) [RT #32688]
Prevents named from aborting with a require assertion failure
on servers with DNS64 enabled. These crashes might occur as a
result of specific queries that are received. (CVE-2012-5688)
[RT #30792 / #30996]
Prevents an assertion failure in named when RPZ and DNS64 are
used together. (CVE-2012-5689) [RT #32141]
See release notes for further features and bug fixes:
https://kb.isc.org/article/AA-00970/0/BIND-9.9.3-P1-Extended-Support-Version-Release-Notes.html
Security: CVE-2013-3919
CVE-2013-2266
CVE-2012-5688
CVE-2012-5689
- address the issue raised by Bob Harold. RRL on recursive servers
applies rate limits after waiting for recursion except on
sub-domains of domains for which the server is authoritative.
- fix the bug reported by Roy Arends in which "slipped" NXDOMAIN
responses had rcode values of 0 (NoError) instead of 3 (NXDOMAIN).
- move reports of RRL drop and slip actions from the "queries"
log category to the "query-errors" category. Because they are not
in the "queres" category, enabling or disabling query logging no
longer affects them.
This is done in a similar manner as the old bind-tools, but uses bind99
instead of bind97 as master port.
Change bind99 to facilitate the bind-tools slave, in a simlar way as was
done for bind97.
Approved by: erwin (maintainer)
The change makes "slip 1;" send only truncated (TC=1) responses.
Without the change, "slip 1;" is the same as the default of "slip 2;".
That default, which alternates truncated with dropped responses
when the rate limit is exceeded, is better for authoritative DNS
servers, because it further reduces the amplification of an attack
from about 1X to about 0.5X.
DNS RRL is not recommended for recursive servers.
Feature safe: yes
Removed the check for regex.h in configure in order
to disable regex syntax checking, as it exposes
BIND to a critical flaw in libregex on some
platforms. [RT #32688]
Security: CVE-2013-2266
working files that should not have been in the patches[1]
Also move to a versioned filename for the patches[2]
Submitted by: Robert Sargent <robtsgt@gmail.com> [1],
Vernon Schryver <vjs@rhyolite.com> [2]
A serious Multiple Zone Response Policy Zone (RPZ2)
Speed Improvement bug has been fixed.
`./configure --enable-rpz-nsip --enable-rpz-nsdname`
is now the default.
Responses affected by the all-per-second parameter
are always dropped. The slip value has no effect on them.
There are improved log messages for responses that are
dropped or "slipped," because they would require an
excessive identical referral.
released version of January 5, 2013.
This also includes performance patches to the BIND9
Response Policy Zones (DNS RPZ), Single Zone Response
Policy Zone (RPZ) Speed Improvement, in the same
patch.
More information: http://ss.vix.su/~vjs/rrlrpz.html
BIND 9 nameservers using the DNS64 IPv6 transition mechanism are
vulnerable to a software defect that allows a crafted query to
crash the server with a REQUIRE assertion failure. Remote
exploitation of this defect can be achieved without extensive
effort, resulting in a denial-of-service (DoS) vector against
affected servers.
Security: 2892a8e2-3d68-11e2-8e01-0800273fe665
CVE-2012-5688
Feature safe: yes
Prevents a crash when queried for a record whose RDATA exceeds
65535 bytes.
Prevents a crash when validating caused by using "Bad cache" data
before it has been initialized.
ISC_QUEUE handling for recursive clients was updated to address
a race condition that could cause a memory leak. This rarely
occurred with UDP clients, but could be a significant problem
for a server handling a steady rate of TCP queries.
A condition has been corrected where improper handling of
zero-length RDATA could cause undesirable behavior, including
termination of the named process.
For more information: https://kb.isc.org/article/AA-00788
in BIND9
High numbers of queries with DNSSEC validation enabled can cause an
assertion failure in named, caused by using a "bad cache" data structure
before it has been initialized.
CVE: CVE-2012-3817
Posting date: 24 July, 2012
from ISC. These patched versions contain a critical bugfix:
Processing of DNS resource records where the rdata field is zero length
may cause various issues for the servers handling them.
Processing of these records may lead to unexpected outcomes. Recursive
servers may crash or disclose some portion of memory to the client.
Secondary servers may crash on restart after transferring a zone
containing these records. Master servers may corrupt zone data if the
zone option "auto-dnssec" is set to "maintain". Other unexpected
problems that are not listed here may also be encountered.
All BIND users are strongly encouraged to upgrade.
the latest from ISC. These versions all contain the following:
Feature Change
* BIND now recognizes the TLSA resource record type, created to
support IETF DANE (DNS-based Authentication of Named Entities)
[RT #28989]
Bug Fix
* The locking strategy around the handling of iterative queries
has been tuned to reduce unnecessary contention in a multi-
threaded environment.
Each version also contains other critical bug fixes.
All BIND users are encouraged to upgrade to these latest versions.
stay as a -devel until it's formally released, which should be soon'ish.
BIND 9.9 includes a number of changes from BIND 9.8 and earlier releases,
including:
NXDOMAIN redirection
Improved startup and reconfiguration time, especially with large
numbers of authoritative zones
New "inline-signing" option, allows named to sign zones completely
transparently, including static zones
Many other new features, especially for DNSSEC
See the CHANGES file for more information on features.
https://kb.isc.org/article/AA-00592