Rather than needing to create intermediate stream variables, we can use
helper functions like `tx_from_hex` instead or access the result
directly, leading both to increased readability and less code.
Passing a list of transactions `txlist` to `create_block` appends
them to the block, hence we don't need to do that manually anymore.
The merkle root calculation can also be removed, since this is done
in the end of the helper.
The previous diff touched most files in ./test/, so bump the headers to
avoid having to touch them again for a bump later.
-BEGIN VERIFY SCRIPT-
./contrib/devtools/copyright_header.py update ./test/
-END VERIFY SCRIPT-
e6998838e5 doc: Add IPv6 address to zmq example (nthumann)
8abe5703a9 test: Add IPv6 test to zmq (nthumann)
ded449b726 zmq: Enable IPv6 on listening socket (nthumann)
Pull request description:
This PR adds support for listening on IPv6 addresses with bitcoinds ZMQ interface, just like the RPC server.
Currently, it is not possible to specify an IPv6 address, as the `ZMQ_IPV6` [socket option](http://api.zeromq.org/master:zmq-setsockopt#toc27) is not set and therefore the ZMQ initialization fails, if one does so. The absence of this option has also been noted [here](https://github.com/bitcoin/bitcoin/issues/15198#issuecomment-617378512).
With this PR one can e.g. set `-zmqpubhashblock=tcp://[::1]:28333` to listen on the IPv6 loopback address.
ACKs for top commit:
laanwj:
Code review ACK e6998838e5
theStack:
Tested ACK e6998838e5🌱
Tree-SHA512: 43c3043d8d5c79794d475926259c1be975b694db4fcc1f7750a9a28e242f0fa1b531735a63ea5777498003aa5834f6243f39742d0f3941f2f37593d0c7890700
It turned out that the "sync up" procedure of repeatedly generating a
block and waiting for a notification once with timeout is too naive in
its current form, as the following scenario could happen:
- generate block A
- receive notification, timeout happens -> repeat procedure
- generate block B
- node publishes block A notification
- receive notification, we receive the one caused by block A
-> sync-up procedure is completed
- node publishes block B
- the actual test starts
- on the first notification reception, one caused by block B is received,
rather than the one actually caused by test code, leading to failure
This change ensures that after each test block generation, we wait for
the notification that is actually caused by that block and ignore others
from possibly earlier blocks.
Co-authored-by: Jon Atack <jon@atack.com>
Speeds up the zmq test roughly by a factor of 2x (~20 sec. instead of
~40 sec.) and also avoids timeouts on the synchronization methods
(sync_mempool() / sync_blocks()) that happened with a slight chance.
This is due to the fact that there is no upper bound on the trickle
relay time, so even the default of 60s is sometimes too low. Fixed by
enabling immediate tx relay on node1.
After connecting the subscriber sockets to the node, there is no
guarantee that the node's zmq publisher interfaces are ready yet, which
means that potentially the first expected notification messages could
get lost and the test fails. Currently this is handled by just waiting
for a short period of time (200ms), which works most of the time but is
still problematic, as in some rare cases the setup time takes much
longer, even in the range of multiple seconds.
The solution in this commit approaches the problem by using a more
robust method of syncing up, originally proposed by instagibbs:
1. Generate a block on the node
2. Try to receive a notification on all subscribers
3. If all subscribers get a message within the timeout (1 second),
we are done, otherwise repeat starting from step 1
The ZMQSubscriber reception methods currently assert that the first
received publisher message has a sequence number of zero. In order to
fix the current test flakiness via "syncing up" to nodes in the setup
phase, we have to cope with the situation that messages get lost and the
first actual received message has a sequence number larger than zero.
