At every site where we've dug into backpressure disconnect findings, it has been
the case that the default values were too small. 8192, so 4x the old default,
suffices at every site to drown out premature disconnects.
With metrics now available for the send buffers regardless of backpressure
overflow policy, this also switches the default from "disconnect" to
"drop_oldest" (for both peers and websockets), meaning that peerings remain
untouched but the oldest queued message simply gets dropped when a new message
is enqueued. With this policy, the number of backpressure overflows is then
simply the count of discarded messages, something that users can tune to see
drop to zero in everyday use. Another benefit is that marginal overflows cause
less message loss than when an entire buffer's worth (plus potentially more
in-flight messages) gets thrown out with a disconnect.
Since a reorg in the Broker library (commit b04195183) that revamped flow
control and that we pulled in with Zeek 5.0, this setting hasn't done
anything. Broker's endpoint::make_subscriber() and
endpoint::make_status_subscriber() take a queue size argument (with a default
value) that simply gets dropped in the eventual subscriber::make() call. See:
b041951835 (diff-5c0d2baa7981caeb6a4080708ddca6ad929746d10c73d66598e46d7c2c03c8deL34-R178)
This implements basic tracking of each peering's current fill level, the maximum
level over a recent time interval (via a new Broker::buffer_stats_reset_interval
tunable, defaulting to 1min), and the number of times a buffer overflows. For
the disconnect policy this is the number of depeerings, but for drop_newest and
drop_oldest it implies the number of messages lost.
This doesn't use "proper" telemetry metrics for a few reasons: this tracking is
Broker-specific, so we need to track each peering via endpoint_ids, while we
want the metrics to use Cluster node name labels, and the latter live in the
script layer. Using broker::endpoint_id directly as keys also means we rely on
their ability to hash in STL containers, which should be fast.
This does not track the buffer levels for Broker "clients" (as opposed to
"peers"), i.e. WebSockets, since we currently don't have a way to name these,
and we don't want to use ephemeral Broker IDs in their telemetry.
To make the stats accessible to the script layer the Broker manager (via a new
helper class that lives in the event_observer) maintains a TableVal mapping
Broker IDs to a new BrokerPeeringStats record. The table's members get updated
every time that table is requested. This minimizes new val instantiation and
allows the script layer to customize the BrokerPeeringStats record by redefing,
updating fields, etc. Since we can't use Zeek vals outside the main thread, this
requires some care so all table updates happen only in the Zeek-side table
updater, PeerBufferState::GetPeeringStatsTable().
Logic to detect this error already existed, but due to enum identifiers
not having a value set, it never triggered before.
Should probably backport this one.
The new Broker API allows us to provide a custom logger to Broker that
pulls previously unattainable context information out of Broker to put
them into broker.log for users of Zeek.
Since Broker log events happen asynchronously, we cache them in a queue
and use a flare to notify Zeek of activity. Furthermore, the Broker
manager now implements the `ProcessFd` function to avoid unnecessary
polling of the new log queue. As a side effect, data stores are polled
less as well.
This exposes Broker's new WebSocket support in Zeek. To enable it,
call `Broker::listen_websocket()`. Zeek will then start listening on
port 9997 for incoming WebSocket connections.
See the Broker documentation for a description of the message format
expected over these WebSocket connections.
Logs that got sent sparsely or burstily would get buffered for long
periods of time since the logic to flush them only does so on the next
log write. In the worst case, a subsequent log write could never happen
and cause a log entry to be indefinitely buffered.
This fix introduces a recurring event/timer to simply flush all pending
logs at frequency of Broker::log_batch_interval.
It may generally be better for our default use-case, as workers may
save a few percent cpu utilization as this policy does not have to
use any polling like the stealing policy does.
This also helps avoid a potential issue with the implementation of
spinlocks used in the work-stealing policy in current CAF versions,
where there's some conditions where lock contention causes a thread
to spin for long periods without relinquishing the cpu to others.
For backward compatibility when reading values, we first check
the ZEEK-prefixed value, and if not set, then check the corresponding
BRO-prefixed value.
* All "Broxygen" usages have been replaced in
code, documentation, filenames, etc.
* Sphinx roles/directives like ":bro:see" are now ":zeek:see"
* The "--broxygen" command-line option is now "--zeexygen"
