Now, during the zero-copy forwarding negotiation, when the requested size is
exact, we are now able to check if it is bigger than the expected one or
not. If it is indeed bigger than expeceted, the zero-copy forwarding is
disabled, the error will be triggered later on the normal sending path.
It is now possible to use a flag during zero-copy forwarding negotiation to
specify the requested size is exact, it means the producer really expect to
receive at least this amount of data.
It can be used by consumer to prepare some processing at this stage, based
on the requested size. For instance, in the H1 mux, it is used to write the
next chunk size.
It is now possible to impose the length to represent the chunk size in the
function used to prepended the chunk size in a buffer (so before the chunk
itself). It is thus possible to reserve a specific space for an unknown
chunk size and padding it with leading '0' to use all the space and avoid
holes.
During zero-copy forwarding negotiation, a pseudo flag was already used to
notify the consummer if the producer is able to use kernel splicing or not. But
this was not extensible. So, now we use a true bitfield to be able to pass flags
during the negotiation. NEGO_FF_FL_* flags may be used now.
Of course, for now, there is only one flags, the kernel splicing support on
producer side (NEGO_FF_FL_MAY_SPLICE).
The cache applet will be refactored to use its own buffer. Thus, for now,
the zero-copy forwarding support is removed and it will be reintrocuded
later.
The HTTP stat applets and all internal functions was adapted to use its own
buffers instead of the channels ones. The CLI part was not refactored yet,
thus there are still some access to channels in the file. But for the HTTP
part, we no longer use the channels at all.
To do so, the HTTP stats applet now uses default .rcv_buf and .snd_buf
callback function. In addition, it sets appctx flags instead of SE ones.
We no longer test the opposite stream-connector to detect aborted partial
post. Applets must not try to access to info ouside their scope. This make
the code more sensitive to changes and it is a common source of bug.
Tests on the sedesc flags at the begining of the I/O handler should be
enough.
Thanks to this patch, it is possible to an applet to directly send data to
the opposite endpoint. To do so, it must implement <fastfwd> appctx callback
function and set SE_FL_MAY_FASTFWD flag.
Everything will be handled by appctx_fastfwd() function. The applet is only
responsible to transfer data. If it sets <to_forward> value, it is used to
limit the amount of data to forward.
This patch introduces the support for the callback function responsible to
produce data via the zero-copy forwarding mechanism. There is no
implementation for now. But <to_forward> field was added in the appctx
structure to let an applet inform how much data it want to forward. It is
not mandatory but it will be used during the zero-copy forwarding
negociation.
We have indroduced flags to deal with end of input, end of stream and errors
at the applet level. With this patch we make the link with the endpoint
descriptor.
In appctx_rcv_buf(), applet flags are converted to SE flags.
There is no shutdown for reads and send with applets. Both are performed
when the appctx is released. So instead of 2 flags, like for
muxes/connections, only one flag is used. But the idea is the same:
acknowledge the event at the applet level.
The appctx state was never really used as a state. It is only used to know
when an applet should be freed on the next wakeup. This can be converted to
a flag and the state can be removed. This is what this patch does.
Till now, we've extended the appctx state to add some flags. However, the
field name is misleading. So a bitfield was added to handle real flags. And
helper functions to manipulate this bitfield were added.
A dedicated function to run applets was introduced, in addition to the old
one, to deal with applets that use their own buffers. The main differnce
here is that this handler does not use channels at all. It performs a
synchronous send before calling the applet and performs a synchronous
receive just after.
No applets are plugged on this handler for now.
There is no tasklet to handle I/O subscriptions for applets, but functions
to deal with receives and sends from the SC layer were added. it meanse a
function to retrieve data from an applet with this synchronous version and a
function to push data to an applet wit this synchronous version.
It is pretty similar to the functions used for muxes but there are some
differences. So for now, we keep them separated.
Zero-copy forwarding is not supported for now. In addition, there is no
subscription mechanism.
In this patch, we add default functions to copy data from a channel to the
<inbuf> buffer of an applet (appctx_rcv_buf) and another on to copy data
from <outbuf> buffer of an applet to a channel (appctx_snd_buf).
These functions are not used for now, but they will be used by applets to
define their <rcv_buf> and <snd_buf> callback functions. Of course, it will
be possible for a specific applet to implement its own functions but these
ones should be good enough for most of applets. HTX and RAW buffers are
supported.
It is the first patch of a series aimed to align applets on connections.
Here, dedicated buffers are added for applets. For now, buffers are
initialized and helpers function to deal with allocation are added. In
addition, flags to report allocation failures or full buffers are also
introduced. <inbuf> will be used to push data to the applet from the stream
and <outbuf> will be used to push data from the applet to the stream.
sc_attach_applet() was changed to be able to fail and callers were updated
accordingly. For now it cannot fail but if this changes, callers will be
prepared to handle errors.
In sc_attach_applet, an untyped pointer (void *) was used to attach a SC on
an applet. There is no reason to not use the right type here. So now a
pointer on an appctx is explicitly used.
Avoid loss of precision when computing K cubic value.
Same issue when computing the congestion window value from cubic increase function
formula with possible integer varaiable wrap around.
Depends on this commit:
MINOR: quic: Code clarifications for QUIC CUBIC (RFC 9438)
Must be backported as far as 2.6.
The first version of our QUIC CUBIC implementation is confusing because relying on
TCP CUBIC linux kernel implementation and with references to RFC 8312 which is
obsoleted by RFC 9438 (August 2023) after our implementation. RFC 8312 is a little
bit hard to understand. RFC 9438 arrived with much more clarifications.
So, RFC 9438 is about "CUBIC for Fast Long-Distance Networks". Our implementation
for QUIC is not very well documented. As it was difficult to reread this
code, this patch adds only some comments at complicated locations and rename
some macros, variables without logic modifications at all.
So, the aim of this patch is to add first some comments and variables/macros renaming
to avoid embedding too much code modifications in the same big patch.
Some code modifications will come to adapt this CUBIC implementation to this new
RFC 9438.
Rename some macros:
CUBIC_BETA -> CUBIC_BETA_SCALED
CUBIC_C -> CUBIC_C_SCALED
CUBIC_BETA_SCALE_SHIFT -> CUBIC_SCALE_FACTOR_SHIFT (this is the scaling factor
which is used only for CUBIC_BETA_SCALED)
CUBIC_DIFF_TIME_LIMIT -> CUBIC_TIME_LIMIT
CUBIC_ONE_SCALED was added (scaled value of 1).
These cubic struct members were renamed:
->tcp_wnd -> ->W_est
->origin_point -> ->W_target
->epoch_start -> ->t_epoch
->remaining_tcp_inc -> remaining_W_est_inc
Local variables to quic_cubic_update() were renamed:
t -> elapsed_time
diff ->t
delta -> W_cubic_t
Add a grahpic curve about the CUBIC Increase function.
Add big copied & pasted RFC 9438 extracts in relation with the 3 different increase
function regions.
Same thing for the fast convergence.
Fix a typo about the reference to QUIC RFC 9002.
Must be backported as far as 2.6 to ease any further modifications to come.
If we were to enable 'ocsp-update' on a certificate that does not have
an OCSP URI, we would exit ssl_sock_load_ocsp with a negative error code
which would raise a misleading error message ("<cert> has an OCSP URI
and OCSP auto-update is set to 'on' ..."). This patch simply fixes the
error message but an error is still raised.
This issue was raised in GitHub #2432.
It can be backported up to branch 2.8.
Fetch will return true if the stream underwent a redispatch according to
"option redispatch" setting upon retries.
Documentation was added, and the "%rc" logformat alternative now mentions
the new fetch to properly emulate the logformat behavior.
This build issued was introduced by this previous commit which is a bugfix:
BUG/MINOR: quic: Wrong ack ranges handling when reaching the limit.
A BUG_ON() referenced <fist> variable in place of <first>.
Must be backported as far as 2.6 as the previous commit.
Acknowledgements ranges are used to build ACK frames. To avoid allocating too
much such objects, a limit was set to 32(QUIC_MAX_ACK_RANGES) by this commit:
MINOR: quic: Do not allocate too much ack ranges
But there is an inversion when removing the oldest range from its tree.
eb64_first() must be used in place of eb64_last(). Note that this patch
only does this modification in addition to rename <last> variable to <first>.
This bug leads such a h2load command to block when a request ends up not
being acknowledged by haproxy even if correctly served:
/opt/nghttp2/build/bin/h2load --alpn-list h3 -t 1 -c 1 -m 1 -n 100 \
https://127.0.0.1/?s=5m
There is a remaining question to be answered. In such a case, haproxy refuses to
reopen the stream, this is a good thing but should not haproxy ackownledge the
request (because correctly parsed again).
Note that to be easily reproduced, this setting had to be applied to the client
network interface:
tc qdisc add dev eth1 root netem delay 100ms 1s loss random
Must be backported as far as 2.6.
As noticed in this thread, some bogus configurations are not always easy
to spot: https://www.mail-archive.com/haproxy@formilux.org/msg44558.html
Here it was about config keywords being used in ACL patterns where strings
were expected, hence they're always valid.
Since we have the diag mode (-dD) we can perform some extra checks when
it's used, and emit them to suggest the user there might be an issue.
Here we detect a few common words (logic such as "and"/"or"/"||" etc),
C++/JS comments mistakenly used to try to isolate final args, and words
that have the exact name of a sample fetch or an ACL keyword. These checks
are only done in diag mode of course.
Final diags were added in 2.4 by commit 5a6926dcf ("MINOR: diag: create
cfgdiag module"), but it's called too late in the startup process,
because when "-c" is passed, the call is not made, while it's its primary
use case. Let's just move the call earlier.
Note that currently the check in this function is limited to verifying
unicity of server cookies in a backend, so it can be backported as far
as 2.4, but there is little value in insisting if it doesn't backport
easily.
Diag warnings were added in 2.4 by commit 7b01a8dbd ("MINOR: global:
define diagnostic mode of execution") but probably due to the split
function that checks for the mode, they did not reuse the emission of
the version string before the first warning, as was brought in 2.2 by
commit bebd21206 ("MINOR: init: report in "haproxy -c" whether there
were warnings or not"). The effet is that diag warnings are emitted
before the version string if there is no other warning nor error. Let's
just proceed like for the two other ones.
This can be backported to 2.4, though this is of very low importance.
Just like for stick-tables, this patch adds a promex module to dump
resolvers metrics. It adds the "resolver" scope and for now, it dumps
folloowing metrics:
* haproxy_resolver_sent
* haproxy_resolver_send_error
* haproxy_resolver_valid
* haproxy_resolver_update
* haproxy_resolver_cname
* haproxy_resolver_cname_error
* haproxy_resolver_any_err
* haproxy_resolver_nx
* haproxy_resolver_timeout
* haproxy_resolver_refused
* haproxy_resolver_other
* haproxy_resolver_invalid
* haproxy_resolver_too_big
* haproxy_resolver_outdated
It is now possible to selectively retrieve extra counters from stats
modules. H1, H2, QUIC and H3 fill_stats() callback functions are updated to
return a specific counter.
The list of modules registered on the stats to expose extra counters is now
public. It is required to export these counters into the Prometheus
exporter.
set-bc-{mark,tos} actions are pretty similar to set-fc-{mark,tos} to set
mark/tos on packets sent from haproxy to server: set-bc-{mark,tos} actions
act on the whole backend/srv connection: from connect() to connection
teardown, thus they may only be used before the connection to the server
is instantiated, meaning that they are only relevant for request-oriented
rules such as tcp-request or http-request rules. For now their use is
limited to content request rules, because tos and mark informations are
stored directly within the stream, thus it is required that the stream
already exists.
stream flags are used in combination with dedicated stream struct members
variables to pass 'tos' and 'mark' informations so that they are correctly
considered during stream connection assignment logic (prior to connecting
to actually connecting to the server)
'tos' and 'mark' fd sockopts are taken into account in conn hash
parameters for connection reuse mechanism.
The documentation was updated accordingly.
In this patch we add the possibility to use sample expression as argument
for set-fc-{mark,tos} actions. To make it backward compatible with
previous behavior, during parsing we first try to parse the value as
as integer (decimal or hex notation), and then fallback to expr parsing
in case of failure.
The documentation was updated accordingly.
This is a complementary patch to "MINOR: tcp-act: Rename "set-{mark,tos}"
to "set-fc-{mark,tos}"", but for the Lua API.
set_mark and set_tos were kept as aliases for set_fc_mark and set_fc_tos
but they were marked as deprecated.
Using this opportunity to reorder set_mark and set_tos by alphabetical
order.
"set-mark" and "set-tos" only alter packets from haproxy to client
(frontend connection). Since we may add support for equivalent keywords
on server side, we rename them with an explicit name to prevent
confusions.
Thus, we rename:
- "set-mark" to "set-fc-mark"
- "set-tos" to "set-fc-tos"
"set-mark" and "set-tos" were kept as aliases (to "set-fc-mark" and
"set-fc-tos" respectively) for now to prevent config breakage, but they
have been marked as deprecated so they can be removed in future version.
Some CPU time is needlessly wasted in conn_calculate_hash(), because all
params are first copied into a temporary buffer before computing the
hash on the whole buffer. Instead, let's leverage the XXH progressive
hash update functions to avoid expensive memcpys.
A major reorganization of QUIC MUX sending has been implemented. Now
data transfer occur over a single QCS buffer. This has improve
performance but at the cost of restrictions on snd_buf. Indeed, buffer
instances are now shared from stream callback snd_buf up to quic-conn
layer.
As such, snd_buf cannot manipulate freely already present data buffer.
In particular, realign has been completely removed by the previous
patches.
This commit reintroduces a partial realign support. This is only done if
the buffer contains only unsent data, via a new MUX function
qcc_realign_stream_txbuf() which is called during snd_buf.
This commit is a direct follow-up on the major rearchitecture of send
buffering. This patch implements the proper handling of connection pool
buffer temporary exhaustion.
The first step is to be able to differentiate a fatal allocation error
from a temporary pool exhaustion. This is done via a new output argument
on qcc_get_stream_txbuf(). For a fatal error, application protocol layer
will schedule the immediate connection closing. For a pool exhaustion,
QCC is flagged with QC_CF_CONN_FULL and stream sending process is
interrupted. QCS instance is also registered in a new list
<qcc.buf_wait_list>.
A new connection buffer can become available when all ACKs are received
for an older buffer. This process is taken in charge by quic-conn layer.
It uses qcc_notify_buf() function to clear QC_CF_CONN_FULL and to wake
up every streams registered on buf_wait_list to resume sending process.
This commit is a direct follow-up on the major rearchitecture of send
buffering. It allows application protocol to react if current QCS
sending buffer space is too small. In this case, the buffer can be
released to the quic-conn layer. This allows to allocate a new QCS
buffer and retry HTX parsing, unless connection buffer pool is already
depleted.
A new function qcc_release_stream_txbuf() serves as API for app protocol
to release the QCS sending buffer. This operation fails if there is
unsent data in it. In this case, MUX has to keep it to finalize transfer
of unsent data to quic-conn layer. QCS is thus flagged with
QC_SF_BLK_MROOM to interrupt snd_buf operation.
When all data are sent to the quic-conn layer, QC_SF_BLK_MROOM is
cleared via qcc_streams_sent_done() and stream layer is woken up to
restart snd_buf.
Note that a new function qcc_stream_can_send() has been defined. It
allows app proto to check if sending is currently blocked for the
current QCS. For now, it checks QC_SF_BLK_MROOM flag. However, it will
be extended to other conditions with the following patches.
The previous commit was a major rework for QUIC MUX sending process.
Following this, this patch cleans up a few elements that remains but can
be removed as they are duplicated.
Of notable changes, offset fields from QCS and QCC are removed. They are
both equivalent to flow control soft offsets.
A new function qcs_prep_bytes() is implemented. Its purpose is to return
the count of prepared data bytes not yet sent. It also replaces
qcs_need_sending().
Previously, QUIC MUX sending was implemented with data transfered along
two different buffer instances per stream.
The first QCS buffer was used for HTX blocks conversion into H3 (or
other application protocol) during snd_buf stream callback. QCS instance
is then registered for sending via qcc_io_cb().
For each sending QCS, data memcpy is performed from the first to a
secondary buffer. A STREAM frame is produced for each QCS based on the
content of their secondary buffer.
This model is useful for QUIC MUX which has a major difference with
other muxes : data must be preserved longer, even after sent to the
lower layer. Data references is shared with quic-conn layer which
implements retransmission and data deletion on ACK reception.
This double buffering stages was the first model implemented and remains
active until today. One of its major drawbacks is that it requires
memcpy invocation for every data transferred between the two buffers.
Another important drawback is that the first buffer was is allocated by
each QCS individually without restriction. On the other hand, secondary
buffers are accounted for the connection. A bottleneck can appear if
secondary buffer pool is exhausted, causing unnecessary haproxy
buffering.
The purpose of this commit is to completely break this model. The first
buffer instance is removed. Now, application protocols will directly
allocate buffer from qc_stream_desc layer. This removes completely the
memcpy invocation.
This commit has a lot of code modifications. The most obvious one is the
removal of <qcs.tx.buf> field. Now, qcc_get_stream_txbuf() returns a
buffer instance from qc_stream_desc layer. qcs_xfer_data() which was
responsible for the memcpy between the two buffers is also completely
removed. Offset fields of QCS and QCC are now incremented directly by
qcc_send_stream(). These values are used as boundary with flow control
real offset to delimit the STREAM frames built.
As this change has a big impact on the code, this commit is only the
first part to fully support single buffer emission. For the moment, some
limitations are reintroduced and will be fixed in the next patches :
* on snd_buf if QCS sent buffer in used has room but not enough for the
application protocol to store its content
* on snd_buf if QCS sent buffer is NULL and allocation cannot succeeds
due to connection pool exhaustion
One final important aspect is that extra care is necessary now in
snd_buf callback. The same buffer instance is referenced by both the
stream and quic-conn layer. As such, some operation such as realign
cannot be done anymore freely.
