A change in 6aba56ae8 (checking whether a rejected RRset was identical
to the data it would have replaced, so that we could still cache a
signature) inadvertently introduced cases where processing of a
response would continue when previously it would have been skipped.
(cherry picked from commit d0fd9cbe3b)
After a reconfiguration the old view can be left without a valid
'rpzs' member, because when the RPZ is not changed during the named
reconfiguration 'rpzs' "migrate" from the old view into the new
view, so when a query resumes it can find that 'qctx->view->rpzs'
is NULL which query_resume() currently doesn't expect to happen if
it's recursing and 'qctx->rpz_st' is not NULL.
Fix the issue by adding a NULL-check. In order to not split the log
message to two different log messages depending on whether
'qctx->view->rpzs' is NULL or not, change the message to not log
the RPZ policy's "version" which is just a runtime counter and is
most likely not very useful for the users.
(cherry picked from commit 3ea2fbc238)
Checking whether the authority section is properly signed should
be left to the validator. Checking in getsection (dns_message_parse)
was way too early and resulted in resolution failures of lookups
that should have otherwise succeeded.
(cherry picked from commit 83159d0a54)
Add a new dns_rdataset_equals() function to check whether two
rdatasets are equal in DNSSEC terms.
When an rdataset being cached is rejected because its trust
level is lower than the existing rdataset, we now check to see
whether the rejected data was identical to the existing data.
This allows us to cache a potentially useful RRSIG when handling
CD=1 queries, while still rejecting RRSIGs that would definitely
have resulted in a validation failure.
(cherry picked from commit 6aba56ae89)
The value returned by http_send_outgoing() is not used anywhere, so we
make it not return anything (void). Probably it is an omission from
older times.
(cherry picked from commit 2adabe835a)
When handling outgoing data, there were a couple of rarely executed
code paths that would not take into account that the callback MUST be
called.
It could lead to potential memory leaks and consequent shutdown hangs.
(cherry picked from commit 8b8f4d500d)
This commit changes the way how the number of active HTTP streams is
calculated and allows it to scale with the values of the maximum
amount of streams per connection, instead of effectively capping at
STREAM_CLIENTS_PER_CONN.
The original limit, which is intended to define the pipelining limit
for TCP/DoT. However, it appeared to be too restrictive for DoH, as it
works quite differently and implements pipelining at protocol level by
the means of multiplexing multiple streams. That renders each stream
to be effectively a separate connection from the point of view of the
rest of the codebase.
(cherry picked from commit a22bc2d7d4)
Previously we would limit the amount of incoming data to process based
solely on the presence of not completed send requests. That worked,
however, it was found to severely degrade performance in certain
cases, as was revealed during extended testing.
Now we switch to keeping track of how much data is in flight (or ready
to be in flight) and limit the amount of processed incoming data when
the amount of in flight data surpasses the given threshold, similarly
to like we do in other transports.
(cherry picked from commit 05e8a50818)
If a deferred validation on data that was originally queried with
CD=1 fails, we now repeat the query, since the zone data may have
changed in the meantime.
(cherry picked from commit 04b1484ed8)
When a query is made with CD=1, we store the result in the
cache marked pending so that it can be validated later, at
which time it will either be accepted as an answer or removed
from the cache as invalid. Deferred validation was not
attempted when there were no cached RRSIGs for DNSKEY and
DS. We now complete the deferred validation in this scenario.
(cherry picked from commit 8b900d1808)
Views use two types of reference counting - regular and weak, and
when there are no more regular references, the view_flushanddetach()
function destroys or detaches some parts of the view, including
'view->zonetable', while other parts are freed by destroy() when
the last weak reference is detached. Since catalog zones use weak
references to attach a view, it's currently possible that during
shutdown catalog zone processing will try to add a new zone into
an otherwise unused view (because it's shutting down) which doesn't
have an attached zonetable any more. This could cause an assertion
failure. Fix this issue by modifying the dns_view_addzone() function
to expect that 'view->zonetable' can be NULL, and in that case just
return ISC_R_SHUTTINGDOWN.
prio_type was being used in the wrong place to optimize cname_and_other.
We have to first exclude and accepted types and we also have to
determine that the record exists before we can check if we are at
a point where a later CNAME cannot appear.
(cherry picked from commit 5e49a9e4ae)
In #1870, the expiration time of ANCIENT records were printed, but
actually the ancient records are very short lived, and the information
carries a little value.
Instead of printing the expiration of ANCIENT records, print the
expiration time of STALE records.
When the mark_header_ancient() helper function was introduced, couple of
places with duplicate (or almost duplicate) code was missed. Add
missing set_ttl() calls before mark_header_ancient(), so the handling of
expiring headers is same in all places.
(concept cherry picked from commit 58179e6a19)
If we know that the header has ZEROTTL set, the server should never send
stale records for it and the TTL should never be anything else than 0.
The comment was already there, but the code was not matching the
comment.
(cherry picked from commit cfee6aa565)
When the header has been marked as ANCIENT, but the ttl hasn't been
reset (this happens in couple of places), the rdataset TTL would be
set to the header timestamp instead to a reasonable TTL value.
Since this header has been already expired (ANCIENT is set), set the
rdataset TTL to 0 and don't reuse this field to print the expiration
time when dumping the cache. Instead of printing the time, we now
just print 'expired (awaiting cleanup'.
(cherry picked from commit 1bbb57f81b)
the search for the deepest known zone cut in the cache could
improperly reject a node containing stale data, even if the
NS rdataset wasn't the data that was stale.
this change also improves the efficiency of the search by
stopping it when both NS and RRSIG(NS) have been found.
(cherry picked from commit 1f095b902c)
Change the names of the node reference counting functions
and add comments to make the mechanism easier to understand:
- new_reference() and decrement_references() are now called
dns__rbtnode_acquire() and dns__rbtnode_release()
respectively; this reflects the fact that they modify both
the internal and external reference counters for a node.
- rbtnode_newref() and rbtnode_decref are now called
rbtnode_erefs_increment() and rbtnode_erefs_decrement(),
to reflect that they only increase and decrease the node's
external reference counters, not internal.
(cherry picked from commit 857225aeb6)
Refactor the pattern in the newref() and decref() functions in rbtdb.c
following the pattern, so it follows the similar pattern we already have
for QPDB.
(cherry picked from commit 9c45de9473)
dns_adb_flushname() calls dns_name_hash() to determine the ADB bucket
number to search for the given name. Meanwhile, all other functions in
lib/dns/adb.c call dns_name_fullhash() for determining the bucket number
instead. This discrepancy causes dns_adb_flushname() to have virtually
no chances of actually removing the given name from the ADB if the
name is longer than 16 bytes (since dns_name_hash() only hashes the
first 16 bytes of the name provided to it) - more specifically, the
probability of success for names longer than 16 bytes is inversely
proportional to the number of ADB buckets in use, i.e. 1:1021 at best.
Fix by using dns_name_fullhash() instead of dns_name_hash() in
dns_adb_flushname(), so that the logic for determining the bucket number
that a given name belongs to is consistent throughout lib/dns/adb.c.
When canceling the last fetch, we also need to stop the fctx_expired
timer from possibly firing between the fctx_shutdown() call and the
fetch being actually destroyed along with the timer. As there are
multiple places where fctx_shutdown() is being called without stopping
the timer, move the fctx_stoptimer() to fctx_shutdown() and cleanup the
explicit usage.
Address Database (ADB) shares the memory for the short lived ADB
objects (finds, fetches, addrinfo) and the long lived ADB
objects (names, entries, namehooks). This could lead to a situation
where the resolver-heavy load would force evict ADB objects from the
database to point where ADB is completely empty, leading to even more
resolver-heavy load.
Make the short lived ADB objects use the other memory context that we
already created for the hashmaps. This makes the ADB overmem condition
to not be triggered by the ongoing resolver fetches.
(cherry picked from commit 05faff6d53)
When the recursive-clients value is too large, the linked lists holding
the fetch contexts can also grow large and since the algorithm to merge
outgoing queries is quadratic, named can get slow.
Replace the linked list with hashtable for faster lookups. This also
allows us to reduce the number of tasks (buckets) in the resolver.
Previously, this function always acquires a node write lock if it
might need node cleanup in case the reference decrements to 0. In
fact, the lock is unnecessary if the reference is larger than 1 and it
can be optimized as an "easy" case. This optimization could even be
"necessary". In some extreme cases, many worker threads could repeat
acquring and releasing the reference on the same node, resulting in
severe lock contention for nothing (as the ref wouldn't decrement to 0
in most cases). This change would prevent noticeable performance
drop like query timeout for such cases.
Co-authored-by: JINMEI Tatuya <jtatuya@infoblox.com>
Co-authored-by: Ondřej Surý <ondrej@isc.org>
(cherry picked from commit 7f4471594d)
Currently, the fetch context will continue running even when the last
fetch (response) has been removed from the context, so named can process
and cache the answer. This can lead to a situation where the number of
outgoing recursing clients exceeds the the configured number for
recursive-clients.
Be more stringent about the recursive-clients limit and shutdown the
fetch context immediately after the last fetch has been canceled from
that particular fetch context.
The last remaining tuning value was RESOLVER_NTASKS and instead of
having variable number of the tasks per-cpu and in named and in
dns_client, set the number of the resolver tasks to 523 (number taken
from dns_client unit) to accomodate most of the recursive-clients
values.
The uv_req union member of struct isc__nm_uvreq contained libuv request
types that we don't use. Turns out that uv_getnameinfo_t is 1000 bytes
big and unnecessarily enlarged the whole structure. Remove all the
unused members from the uv_req union.
After removing sockaddr_unix from isc_sockaddr, we can also remove
sockaddr_storage and reduce the isc_sockaddr size from 152 bytes to just
48 bytes needed to hold IPv6 addresses.
(cherry picked from commit 2367b6a2e1)
We started using isc_nm_bad_request() more actively throughout
codebase. In the case of HTTP/2 it can lead to a large count of
useless "Bad Request" messages in the BIND log, as often we attempt to
send such request over effectively finished HTTP/2 sessions.
This commit fixes that.
(cherry picked from commit 937b5f8349)
This commit introduces manual read timer control as used by StreamDNS
and its underlying transports. Before that, DoH code would rely on the
timer control provided by TCP, which would reset the timer any time
some data arrived. Now, the timer is restarted only when a full DNS
message is processed in line with other DNS transports.
That change is required because we should not stop the timer when
reading from the network is paused due to throttling. We need a way to
drop timed-out clients, particularly those who refuse to read the data
we send.
(cherry picked from commit 609a41517b)
This commit adds logic to make code better protected against clients
that send valid HTTP/2 data that is useless from a DNS server
perspective.
Firstly, it adds logic that protects against clients who send too
little useful (=DNS) data. We achieve that by adding a check that
eventually detects such clients with a nonfavorable useful to
processed data ratio after the initial grace period. The grace period
is limited to processing 128 KiB of data, which should be enough for
sending the largest possible DNS message in a GET request and then
some. This is the main safety belt that would detect even flooding
clients that initially behave well in order to fool the checks server.
Secondly, in addition to the above, we introduce additional checks to
detect outright misbehaving clients earlier:
The code will treat clients that open too many streams (50) without
sending any data for processing as flooding ones; The clients that
managed to send 1.5 KiB of data without opening a single stream or
submitting at least some DNS data will be treated as flooding ones.
Of course, the behaviour described above is nothing else but
heuristical checks, so they can never be perfect. At the same time,
they should be reasonable enough not to drop any valid clients,
realatively easy to implement, and have negligible computational
overhead.
(cherry picked from commit 3425e4b1d0)
Initially, our DNS-over-HTTP(S) implementation would try to process as
much incoming data from the network as possible. However, that might
be undesirable as we might create too many streams (each effectively
backed by a ns_client_t object). That is too forgiving as it might
overwhelm the server and trash its memory allocator, causing high CPU
and memory usage.
Instead of doing that, we resort to processing incoming data using a
chunk-by-chunk processing strategy. That is, we split data into small
chunks (currently 256 bytes) and process each of them
asynchronously. However, we can process more than one chunk at
once (up to 4 currently), given that the number of HTTP/2 streams has
not increased while processing a chunk.
That alone is not enough, though. In addition to the above, we should
limit the number of active streams: these streams for which we have
received a request and started processing it (the ones for which a
read callback was called), as it is perfectly fine to have more opened
streams than active ones. In the case we have reached or surpassed the
limit of active streams, we stop reading AND processing the data from
the remote peer. The number of active streams is effectively decreased
only when responses associated with the active streams are sent to the
remote peer.
Overall, this strategy is very similar to the one used for other
stream-based DNS transports like TCP and TLS.
(cherry picked from commit 9846f395ad)
This commit adds isc__nm_async_run() which is very similar to
isc_async_run() in newer versions of BIND: it allows calling a
callback asynchronously.
Potentially, it can be used to replace some other async operations in
other networking code, in particular the delayed I/O calls in TLS a
TCP DNS transports to name a few and remove quiet a lot of code, but
it we are unlikely to do that for the strictly maintenance only
branch, so it is protected with DoH-related #ifdefs.
It is implemented in a "universal" way mainly because doing it in the
specific code requires the same amount of code and is not simpler.
This commit adds a manual TLS read timer control mode which is
supposed to override automatic resetting of the timer when any data is
received.
It both depends and complements similar functionality in TCP.
This commit adds a manual TCP read timer control mode which is
supposed to override automatic resetting of the timer when any data is
received. That can be accomplished by
`isc__nmhandle_set_manual_timer()`.
This functionality is supposed to be used by multilevel networking
transports which require finer grained control over the read
timer (TLS Stream, DoH).
The commit is essentially an implementation of the functionality from
newer versions of BIND.
When answering queries, don't add data to the additional section if
the answer has more than 13 names in the RDATA. This limits the
number of lookups into the database(s) during a single client query,
reducing query processing load.
Also, don't append any additional data to type=ANY queries. The
answer to ANY is already big enough.
(cherry picked from commit a1982cf1bb)
When isc_rwlock_trylock() fails to get a read lock because another
writer was faster, it should wake up other waiting writers in case
there are no other readers, but the current code forgets about
the currently active writer when evaluating 'cntflag'.
Unset the WRITER_ACTIVE bit in 'cntflag' before checking to see if
there are other readers, otherwise the waiting writers, if they exist,
might not wake up.
Only call eatline() to skip to the next line if we're not
already at the end of a line when parsing an unknown directive.
We were accidentally skipping the next line when there was only
a single unknown directive on the current line.
(cherry picked from commit eb78ad2080)
Commit af7db89513 as part of #4141 was
supposed to apply the 'max-recursion-queries' quota to validator
queries, but the counter was never actually passed on to
dns_resolver_createfetch(). This has been fixed, and the global query
counter ('max-query-count', per client request) is now also added.
(cherry picked from commit 5b1ae4a948)
Upstream code doesn't do regular releases, so we need to regularly
sync the code from the upstream repository. This is synchronization up
to the commit f8d0513 from Jan 29, 2024.
(cherry picked from commit d14a76e115)
While implementing the global limit 'max-query-count', initially I
thought adding the variable to the resolver structure. But the limit
is per client request so it was moved to the view structure (and
counter in ns_query structure). However, I forgot to remove the
variable from the resolver structure again. This commit fixes that.
(cherry picked from commit 397ca34e34)
This global limit is not reset on query restarts and is a hard limit
for any client request.
Note: This commit has been significantly modified because of many
merge conflicts due to the dns_resolver_createfetch api changes.
(cherry picked from commit 16b3bd1cc7)
Add another option to configure how many outgoing queries per
client request is allowed. The existing 'max-recursion-queries' is
per restart, this one is a global limit.
(cherry picked from commit bbc16cc8e6)
The root cause is the fix for CVE-2024-0760 (part 3), which resets
the TCP connection on a failed send. Specifically commit
4b7c6138 stops reading on the socket
because the TCP connection is throttling.
When the tcpdns_send_cb callback thinks about restarting reading
on the socket, this fails because the socket is a client socket.
And nsupdate is a client and is using the same netmgr code.
This commit removes the requirement that the socket must be a server
socket, allowing reading on the socket again after being throttled.
(manually picked from commit aa24b77d8b)
Previously, the update policy rules check was moved earlier in the
sequence, and the keep rule match pointers were kept to maintain the
ability to verify maximum records by type.
However, these pointers can become invalid if server reloading
or reconfiguration occurs before update completion. To prevent
this issue, extract the maximum records by type value immediately
during processing and only keep the copy of the values instead of the
full ssurule.
(cherry picked from commit 44a54a29d8)
DNS_LOGMODULE_RBTDB was simply inappropriate, and this
log message is actually dependent on db implementation
details, so DNS_LOGMODULE_DB would be the best choice.
(cherry picked from commit b0309ee631)
Evan Hunt
Aram Sargsyan
Mark Andrews
Artem Boldariev
Ondřej Surý
Michał Kępień
Andoni Duarte Pintado
JINMEI TatuyaandOndřej Surý
Artem BoldarievandAndoni Duarte Pintado
Matthijs Mekking