To break a potential read/write deadlock, gossipsub uses an unbounded
queue for writes - when peers are too slow to process this queue, it may
end up growing without bounds causing high memory usage.
Here, we introduce a maximum write queue length after which the peer is
disconnected - the queue is generous enough that any "normal" usage
should be fine - writes that are `await`:ed are not affected, only
writes that are launched in an `asyncSpawn` task or similar.
* avoid unnecessary copy of message when there are no send observers
* release message memory earlier in gossipsub
* simplify pubsubpeer logging
* add helper to read EOF marker after closing stream (else stream stay
alive until timeout/reset)
* don't assert on empty channel message
* don't loop when writing to chronos (no need)
When messages can't be sent to peer, we try to establish a send
connection - this causes messages to stack up as more and more unsent
messages are blocked on the dial lock.
* remove dial lock
* run reconnection loop in background task
* channel close race and deadlock fixes
* remove send lock, write chunks in one go
* push some of half-closed implementation to BufferStream
* fix some hangs where LPChannel readers and writers would not always
wake up
* simplify lazy channels
* fix close happening more than once in some orderings
* reenable connection tracking tests
* close channels first on mplex close such that consumers can read bytes
A notable difference is that BufferedStream is no longer considered EOF
until someone has actually read the EOF marker.
* docs, simplification
* add peer lifecycle events
* rework peer events to not use connection events
* don't use result in pubsub and switch init
* wip
* use ordered hashes and remove logscope
* logging
* add missing test
* small fixes
* remove almost-empty types module
* lock when writing message (that's the only place the lock matters, and
only when the message is > max msg size)
* logging updates (log in consistent order, makes reading logs easier)
* raise EOF from readExactly only if no bytes have been read (to signal
that _no_ bytes were lost)
This change modifies how the backpressure algorithm in bufferstream
works - in particular, instead of working byte-by-byte, it will now work
seq-by-seq.
When data arrives, it usually does so in packets - in the current
bufferstream, the packet is read then split into bytes which are fed one
by one to the bufferstream. On the reading side, the bytes are popped of
the bufferstream, again byte by byte, to satisfy `readOnce` requests -
this introduces a lot of synchronization traffic because the checks for
full buffer and for async event handling must be done for every byte.
In this PR, a queue of length 1 is used instead - this means there will
at most exist one "packet" in `pushTo`, one in the queue and one in the
slush buffer that is used to store incomplete reads.
* avoid byte-by-byte copy to buffer, with synchronization in-between
* reuse AsyncQueue synchronization logic instead of rolling own
* avoid writeHandler callback - implement `write` method instead
* simplify EOF signalling by only setting EOF flag in queue reader (and
reset)
* remove BufferStream pipes (unused)
* fixes drainBuffer deadlock when drain is called from within read loop
and thus blocks draining
* fix lpchannel init order
if the connection is already closed (because the remote closes during
identfiy for example), an exception would be raised which would leave
the connection in limbo, beacuse it would not go through the rest of
internalConnect.
Also, if the connection is already closed, the disconnect event would be
scheduled before the connect event :/
* mcache fixes
* remove timed cache - the window shifting already removes old messages
* ref -> object
* avoid unnecessary allocations with `[]` operator
* simplify init
* fix several gossipsub/floodsub issues
* floodsub, gossipsub: don't rebroadcast messages that fail validation
(!)
* floodsub, gossipsub: don't crash when unsubscribing from unknown
topics (!)
* gossipsub: don't send message to peers that are not interested in the
topic, when messages don't share topic list
* floodsub: don't repeat all messages for each message when
rebroadcasting
* floodsub: allow sending empty data
* floodsub: fix inefficient unsubscribe
* sync floodsub/gossipsub logging
* gossipsub: include incoming messages in mcache (!)
* gossipsub: don't rebroadcast already-seen messages (!)
* pubsubpeer: remove incoming/outgoing seen caches - these are already
handled in gossipsub, floodsub and will cause trouble when peers try to
resubscribe / regraft topics (because control messages will have same
digest)
* timedcache: reimplement without timers (fixes timer leaks and extreme
inefficiency due to per-message closures, futures etc)
* timedcache: ref -> obj
when identify is run on incoming connections, the connmanager tables are
updated too late for incoming connections to properly be handled
this is a quickfix that will eventually need cleaning up
* remove send lock
When mplex receives data it will block until a reader has processed the
data. Thus, when a large message is received, such as a gossipsub
subscription table, all of mplex will be blocked until all reading is
finished.
However, if at the same time a `dial` to establish a gossipsub send
connection is ongoing, that `dial` will be blocked because mplex is no
longer reading data - specifically, it might indeed be the connection
that's processing the previous data that is waiting for a send
connection.
There are other problems with the current code:
* If an exception is raised, it is not necessarily raised for the same
connection as `p.sendConn`, so resetting `p.sendConn` in the exception
handling is wrong
* `p.isConnected` is checked before taking the lock - thus, if it
returns false, a new dial will be started. If a new task enters `send`
before dial is finished, it will also determine `p.isConnected` is
false, then get stuck on the lock - when the previous task finishes and
releases the lock, the new task will _also_ dial and thus reset
`p.sendConn` causing a leak.
* prefer existing connection
simplifies flow
Giovanni Petrantoni
Dean Eigenmann
Mamy Ratsimbazafy
Giovanni Petrantoni
Jacek Sieka
Dmitriy Ryajov
Eugene Kabanov
Oskar Thorén
Giovanni Petrantoni
Zahary Karadjov