Refactor Test Framework Architecture: Responsibility Segregation, Configuration, and Clarity #733

moisesPompilio · 2025-12-16T01:50:06Z

Description and Notes

The test framework for Floresta functional tests has been restructured to better segregate responsibilities and simplify the overall architecture. Previously, logic for node instantiation, configuration, and RPC handling was scattered across multiple code locations, making maintenance and understanding the code flow difficult.

Key architectural changes:

Daemon classes (BitcoinDaemon, FlorestaDaemon, UtreexoDaemon) handle node instantiation, startup, and are solely responsible for managing all relevant configuration flags (network, RPC, P2P, Electrum, etc.), including connection URLs and port assignments.
RPC classes (BitcoinRPC, FlorestaRPC, UtreexoRPC) are now focused exclusively on RPC communications. Shared RPC functionality common to all node types was centralized in the BaseRPC class to reduce duplication and ease maintenance.
The Node class orchestrates node instantiation, assembling Daemon and RPC as needed, and now handles its own internal ElectrumClient initialization for easier Electrum interactions inside tests.

New configuration classes (ConfigP2P, ConfigRPC, ConfigElectrum, ConfigTls) provide structured and customizable handling of different node aspects. For example, custom P2P configuration is now as simple as providing a specific configuration instance, improving test readability and extensibility.

Additional improvements:

API for adding nodes has been clarified, with explicit functions for default nodes, nodes with extra arguments, and nodes with Electrum/TLS support; making test intentions easier to understand at a glance.
The test tls-fail.py was removed as it did not test Floresta functionality, but only ElectrumClient connection handling, which is out of scope for this test suite.

A significant bug was addressed by centrally setting node ports in the configuration stage, removing the need to wait for nodes to start in order to detect port usage. P2P, RPC, and Electrum ports are now set in advance and known to all test framework components, eliminating port detection issues.

How to verify the changes you have done?

Review the new class structure and responsibilities: Daemon classes for startup and config, RPC classes for RPC methods, and Node managing component orchestration and Electrum client usage.
Add nodes to tests using new, clearer APIs for default, custom, or Electrum/TLS-enabled nodes, and observe improved code clarity and flexibility.
Validate that configuration overrides (e.g., custom P2P or RPC settings) work as expected when passing specialized configuration instances to nodes.
Ensure that node startup and port assignment are predictable, with all necessary ports set centrally in configuration, and that no dynamic or post-hoc port detection is required.
Confirm that the removed test (tls-fail.py) is no longer present.
Observe centralized, shared RPC behavior via the BaseRPC class, reducing repetition and simplifying common method updates.

Davidson-Souza · 2025-12-29T16:52:01Z

This needs rebase

luisschwab · 2025-12-29T16:55:41Z

Is the functional test move to Rust still a go? I'm starting to work on halfin, which will be a runner for bitcoin-related binaries, inspired on electrsd. It will be possible to run bitcoind, electrsd, utreexod and florestad.

moisesPompilio · 2025-12-29T17:02:46Z

Is the functional test move to Rust still a go? I'm starting to work on halfin, which will be a runner for bitcoin-related binaries, inspired on electrsd. It will be possible to run bitcoind, electrsd, utreexod and florestad.

For now I'm focusing on improving the existing Python tests. The Rust conversion doesn't have much approval at the moment, so it'll be on hold for now.

moisesPompilio · 2025-12-29T17:31:50Z

This needs rebase

Rebased

Davidson-Souza

partial review, will continue tomorrow (TM)

tests/example/electrum.py

tests/floresta-cli/addnode-v1.py

tests/floresta-cli/addnode-v2.py

tests/floresta-cli/getmemoryinfo.py

tests/test_framework/__init__.py

…g Improvements, and Test Restart Refactoring e87a4d2 feat(functional): validate daemon startup and log errors on failure (moisesPomilio) 26804ae fix(test): improve port detection logic and reduce timeout (moisesPomilio) 60fe445 refactor(test): simplify restart test to use a single node (moisesPomilio) Pull request description: ### Description and Notes A validation step was added after node startup to ensure the daemon actually started successfully. After executing the startup command, the workflow waits for 1 second to verify if the node is running. If the daemon fails to start(such as due to misconfigured flags or command errors) the process is stopped, a relevant error message is logged (including the output of the command and the affected node), and an error is returned. This provides much clearer feedback for developers writing or running tests and makes it easier to detect incorrect flags or startup commands. The port detection logic was also improved. Previously, the test only read the last line of the log within a specified timeout, which could cause failures if the output timing was off, resulting in missed port detection and unstable CI runs. Now, all log lines are read and checked, significantly reducing the chance of missing the port assignment. Since port information is always printed at the beginning of the logs, this change does not impact performance. Note: This is not a definitive solution to the architectural issue behind port detection; a more comprehensive fix is planned for PR #733. Additionally, the restart test was refactored for better clarity and relevance. Instead of starting two nodes, stopping them, and checking their data directories (which did not accurately reflect a restart scenario), the test now starts a node, stops it, and then starts it again, verifying that restart works correctly and that stopping the node does not corrupt files. ### How to verify the changes you have done? - Start a node with both valid and intentionally invalid flags or commands. Confirm that, if the daemon fails to start, a clear log error is shown with the command output and node information. - Run the functional test suite and observe improved reliability in port detection; verify that logs are parsed correctly and the port is assigned even with slight output timing differences. - Review the restart test logic: ensure it starts a node, stops it, and starts the same node again, checking for correct restart behavior and file integrity. ACKs for top commit: jaoleal: ACK e87a4d2 j-moreno-c-r: ACK e87a4d2 Davidson-Souza: ACK e87a4d2 Tree-SHA512: 3f266552b84bf0df0a1decfd828c8600f90afaa2d0c070ffe4e6e0b7bb09cdcd7f661cdeadfd03b1c68d6fb036b2b519da28798a8e6d238deff12a368abdf800

Davidson-Souza · 2026-01-07T14:51:21Z

Needs rebase

moisesPompilio · 2026-01-07T21:19:48Z

Rebased

moisesPompilio · 2026-01-08T21:59:36Z

The adjustments I was making to the addnode-v1 and addnode-v2 test part I moved to PR #767, the reason being it's already going beyond the scope of this PR's purpose.

Davidson-Souza · 2026-01-21T15:56:34Z

Needs rebase

moisesPompilio · 2026-01-21T16:09:27Z

Rebased

Davidson-Souza

I really like the new organization. The tests are much easier to reason about, and I think it will really pay off as we go for more complex scenarios.

Just a few nits from me

tests/test_framework/daemon/bitcoin.py

tests/test_framework/daemon/floresta.py

tests/test_framework/rpc/base.py

tests/test_framework/rpc/bitcoin.py

tests/test_framework/rpc/base.py

Davidson-Souza · 2026-01-21T22:57:45Z

tests/test_framework/__init__.py


    @staticmethod
-    def get_available_random_port(start: int, end: int = 65535):
+    def get_available_random_port_by_range(start: int, end: int):


I still got a transient error here. It happens like once every 30 executions on my machine (yes, I ran this hundreds of times, looking very good!). If the two tests are executed at the same time, the rng is seeded with system clock, so it will return the same port. And since they run at the same time, no one had the chance to bind to the port, so the test in 395 won't catch it.

We can easily solve it by seeding the rng with the test name. Perhaps testname + system time?

Actually, looking at the logs it's more like once every 5 runs on a faster-ish machine

Nice catch! I had already seen this error but didn't investigate it further. I made the fix and on my machine the error apparently stopped happening.

I ran 128dd59 and still breaks. Wft is wrong with this RNG?

Perhaps run_node could re-randomize the port each attempt? Fyi I run this:

while true; do just test-functional-run || break; done

im always receiving a timeout in this branch

Perhaps run_node could re-randomize the port each attempt? Fyi I run this:

I modified the architecture to move the default config creation to the node side. The node can now also have a static_value variable to indicate that values should be static. So when the node is in the initialization process and static_value is false, it changes its config values if there's an error. When the config is successful, it sets static_value to true to prevent an already initialized node from having its values modified if it gets restarted.

moisesPompilio · 2026-01-26T21:54:19Z

f8c5862 I modified the architecture so that a node using default configs can have its config modified during initialization. This happens if the node fails to start, allowing it to change the config, especially the port, which solves the port conflict issue identified in this PR. However, after it's initialized, it's marked as static_value = True to prevent its config from being modified when restarted.

moisesPompilio · 2026-01-26T21:55:45Z

Just to notate here publicly instead of only keeping this between moises and I.

Im having this strange behavior that floresta just dies, this is from the addnodev2 test. I tried changing paths like he suggested but it made no difference

Is this happening when he restarts bitcoind or does this always happen when the test starts?

Davidson-Souza · 2026-01-28T15:19:54Z

With the following diff things got way more stable. It still breaks sometimes, but way less. Perhaps we should go ahead with this one and investigate a little further

diff --git a/crates/floresta-wire/src/p2p_wire/node/conn.rs b/crates/floresta-wire/src/p2p_wire/node/conn.rs
index 25c40f6..de47ab2 100644
--- a/crates/floresta-wire/src/p2p_wire/node/conn.rs
+++ b/crates/floresta-wire/src/p2p_wire/node/conn.rs
@@ -561,6 +561,9 @@ where
         &mut self,
         required_service: ServiceFlags,
     ) -> Result<(), WireError> {
+        // try to connect with manually added peers
+        self.maybe_open_connection_with_added_peers()?;
+
         // If the user passes in a `--connect` cli argument, we only connect with
         // that particular peer.
         if self.fixed_peer.is_some() && !self.peers.is_empty() {
@@ -573,9 +576,6 @@ where
         let needs_utreexo = required_service.has(service_flags::UTREEXO.into());
         self.maybe_use_hardcoded_addresses(needs_utreexo);
 
-        // try to connect with manually added peers
-        self.maybe_open_connection_with_added_peers()?;
-
         let connection_kind = ConnectionKind::Regular(required_service);
         if self.peers.len() < T::MAX_OUTGOING_PEERS {
             self.create_connection(connection_kind)?;
diff --git a/tests/floresta-cli/getbestblockhash.py b/tests/floresta-cli/getbestblockhash.py
index b31f819..0886679 100644
--- a/tests/floresta-cli/getbestblockhash.py
+++ b/tests/floresta-cli/getbestblockhash.py
@@ -57,9 +57,7 @@ class GetBestblockhashTest(FlorestaTestFramework):
 
         self.log("=== Connect floresta to utreexod")
         utreexod_url = self.utreexod.p2p_url
-        self.florestad.rpc.addnode(
-            node=utreexod_url, command="onetry", v2transport=False
-        )
+        self.florestad.rpc.addnode(node=utreexod_url, command="add", v2transport=False)
 
         self.log("=== Waiting for floresta to connect to utreexod...")
         self.wait_for_peers_connections(self.florestad, self.utreexod)
diff --git a/tests/floresta-cli/getblockcount.py b/tests/floresta-cli/getblockcount.py
index b72f3ad..b838103 100644
--- a/tests/floresta-cli/getblockcount.py
+++ b/tests/floresta-cli/getblockcount.py
@@ -71,14 +71,14 @@ class GetBlockCountTest(FlorestaTestFramework):
 
         self.log("=== Connect floresta to utreexod")
         utreexod_url = self.utreexod.p2p_url
-        self.florestad.rpc.addnode(utreexod_url, command="onetry", v2transport=False)
+        self.florestad.rpc.addnode(utreexod_url, command="add", v2transport=False)
 
         self.log("=== Waiting for floresta to connect to utreexod...")
         self.wait_for_peers_connections(self.florestad, self.utreexod)
 
         self.log("=== Connect bitcoind to utreexod")
         bitcoind_url = self.bitcoind.p2p_url
-        self.bitcoind.rpc.addnode(utreexod_url, command="onetry", v2transport=False)
+        self.bitcoind.rpc.addnode(utreexod_url, command="add", v2transport=False)
 
         self.log("=== Waiting for bitcoind to connect to utreexod...")
         self.wait_for_peers_connections(self.bitcoind, self.utreexod)
diff --git a/tests/floresta-cli/getblockhash.py b/tests/floresta-cli/getblockhash.py
index a7debf8..dea6efc 100644
--- a/tests/floresta-cli/getblockhash.py
+++ b/tests/floresta-cli/getblockhash.py
@@ -52,13 +52,13 @@ class GetBlockhashTest(FlorestaTestFramework):
 
         self.log("=== Connect floresta to utreexod")
         utreexod_url = self.utreexod.p2p_url
-        self.florestad.rpc.addnode(utreexod_url, command="onetry", v2transport=False)
+        self.florestad.rpc.addnode(utreexod_url, command="add", v2transport=False)
 
         self.log("=== Waiting for floresta to connect to utreexod...")
         self.wait_for_peers_connections(self.florestad, self.utreexod)
 
         self.log("=== Connect bitcoind to utreexod")
-        self.bitcoind.rpc.addnode(utreexod_url, command="onetry", v2transport=False)
+        self.bitcoind.rpc.addnode(utreexod_url, command="add", v2transport=False)
 
         self.log("=== Waiting for bitcoind to connect to utreexod...")
         self.wait_for_peers_connections(self.bitcoind, self.utreexod)
diff --git a/tests/floresta-cli/gettxout.py b/tests/floresta-cli/gettxout.py
index de63d93..8f23c4c 100644
--- a/tests/floresta-cli/gettxout.py
+++ b/tests/floresta-cli/gettxout.py
@@ -70,13 +70,13 @@ class GetTxoutTest(FlorestaTestFramework):
 
         self.log("=== Connect floresta to utreexod")
         utreexod_url = self.utreexod.p2p_url
-        self.florestad.rpc.addnode(utreexod_url, command="onetry", v2transport=False)
+        self.florestad.rpc.addnode(utreexod_url, command="add", v2transport=False)
 
         self.log("=== Waiting for floresta to connect to utreexod...")
         self.wait_for_peers_connections(self.florestad, self.utreexod)
 
         self.log("=== Connect bitcoind to utreexod")
-        self.bitcoind.rpc.addnode(utreexod_url, command="onetry", v2transport=False)
+        self.bitcoind.rpc.addnode(utreexod_url, command="add", v2transport=False)
 
         self.log("=== Waiting for bitcoind to connect to utreexod...")
         self.wait_for_peers_connections(self.bitcoind, self.utreexod)
diff --git a/tests/floresta-cli/ping.py b/tests/floresta-cli/ping.py
index d8ac84f..b78f99a 100644
--- a/tests/floresta-cli/ping.py
+++ b/tests/floresta-cli/ping.py
@@ -21,7 +21,7 @@ class PingTest(FlorestaTestFramework):
         self.run_node(self.bitcoind)
 
         # Connect floresta to bitcoind
-        self.florestad.rpc.addnode(self.bitcoind.p2p_url, "onetry")
+        self.florestad.rpc.addnode(self.bitcoind.p2p_url, "add")
 
         time.sleep(1)
 
diff --git a/tests/florestad/reorg-chain.py b/tests/florestad/reorg-chain.py
index 2162c81..d37e917 100644
--- a/tests/florestad/reorg-chain.py
+++ b/tests/florestad/reorg-chain.py
@@ -41,7 +41,7 @@ class ChainReorgTest(FlorestaTestFramework):
 
         self.log("=== Connect floresta to utreexod")
         self.florestad.rpc.addnode(
-            self.utreexod.p2p_url, command="onetry", v2transport=False
+            self.utreexod.p2p_url, command="add", v2transport=False
         )
 
         self.log("=== Waiting for floresta to connect to utreexod.rpc...")

moisesPompilio · 2026-02-04T16:33:20Z

7a6ac6e I added the connect_for_peers method to centralize peer connection logic. When I did that I noticed the __init__.py file of the test_framework exceeded 1000 lines, which caused pylint to raise a file extension/size error, so I moved the Node class into a separate node.py file and moved some of the utilities that were in the test_framework into util.py. I also ran the tests to find where the random failures occurred. I found that the gettxout.py test had an error that happened when IBD was not fully completed, so I added a better node sync rule. In addition, I noticed that sometimes nodes started but were not yet ready to receive RPC requests or Electrum server requests, so when starting the node it now also checks whether the node is ready to receive requests. I also observed that sometimes nodes take a long time to connect to each other, so in wait_for_peers_connections I increased the wait time to check whether a node is connected; during the first 10 attempts the interval between requests is 1 second and after that it becomes 2 seconds, to give a node that is taking longer more time to connect without causing excessive processing when responding to requests. I also applied the suggestions proposed here

Davidson-Souza · 2026-02-04T21:23:43Z

It's way more reliable now. I did get a crash, but after hundreds of runs. I'll review ASAP, but I think it's pretty much ready

Davidson-Souza · 2026-02-04T22:00:52Z

tests/floresta-cli/disconnectnode.py

        )
        # Call `disconnectnode` with an invalid `node_address` (wrong port).
-        node_address = f"127.0.0.1:{self.bitcoind.get_port('p2p') + 1}"
+        node_address = f"{self.bitcoind.p2p_url}1"


01eaa31

This could generate invalid ports. The test needs a valid IP/port that doesn't have any associated peer.

I reverted it to the way it was before; for this part I only split the new information we have to obtain the port and the IP separately.

Davidson-Souza · 2026-02-04T22:02:10Z

tests/floresta-cli/disconnectnode.py

        )
        # Call `disconnectnode` with an invalid `node_address` (wrong IP address: 127.0.0.2).
-        node_address = f"127.0.0.2:{self.bitcoind.get_port('p2p')}"
+        node_address = f"1{self.bitcoind.p2p_url}"


01eaa31

You can keep the same ip as in the original test. This IP is invalid (the test requires a valid ip that doesn't have an associated peer)

I reverted to the original state to keep the test as it was.

Davidson-Souza · 2026-02-04T22:02:27Z

tests/floresta-cli/disconnectnode.py

        )
        # Call `disconnectnode` with an invalid `node_address` (wrong IP address).
-        node_address = f"127.0.0:{self.bitcoind.get_port('p2p')}"
+        node_address = f"1.{self.bitcoind.p2p_url}"


01eaa31

Here too

I reverted to the original state to keep the test as it was.

tests/test_framework/daemon/base.py

tests/test_framework/__init__.py

This refactor introduces a new architecture for the test framework, separating responsibilities across classes to improve maintainability and clarity: - Daemon classes (`BitcoinDaemon`, `FlorestaDaemon`, `UtreexoDaemon`) are responsible for instantiating and configuring nodes, managing the necessary flags and settings to run a node. - RPC classes (`BitcoinRPC`, `FlorestaRPC`, `UtreexoRPC`) handle only RPC-related functionality, such as managing connections and performing requests. - The `Node` class is now responsible for controlling the instantiation of a node by assembling the required components (Daemon and RPC) and managing their interactions. This reorganization also fixes a bug in port handling by explicitly defining RPC and P2P ports in the configurations, which are passed directly to the `Node` class. Additional changes: - Introduced `ConfigP2P`, `ConfigRPC`, `ConfigElectrum`, and `ConfigTls` classes for structured configuration handling. - Simplified `BaseDaemon` class and `BaseRPC` class by removing the metaclass and enforcing method implementations through abstract methods. - Added the ability to update the node's configurations (RPC, P2P, Electrum) dynamically when `static_values` is set to False. This ensures that if a node fails to start due to issues like port conflicts, the configurations can be regenerated and retried automatically.

Moved RPC methods that are common across `UtreexoRPC`, `FlorestaRPC`, and `BitcoinRPC` to the `BaseRPC` class to reduce code duplication and centralize shared functionality.

Removed `TestSslFailInitialization` from `florestad/tls-fail.py` as it does not test Floresta's behavior. The test only validates `ElectrumClient` connection handling, which is unrelated to Floresta's functionality.

The `Node` class now initializes an `ElectrumClient` instance internally, simplifying its usage and removing the need to instantiate it externally.

- Moved Node class and NodeType enum to a new file `node.py` for better organization. - Introduced Utility class in `util.py` for common helper functions, including directory and port management. - Updated all test files to import Node and Utility from their new locations. - Replaced direct calls to utility functions in the test framework with calls to the new Utility class methods.

…ork` and `Node` - Introduced the `connect_nodes` method in `FlorestaTestFramework`: - Responsible for connecting two peers to each other. - Verifies if the peers are successfully connected. - Added the `connect_node` method to the `Node` class: - Allows the current node to connect to a specified peer. - Improved the `wait_for_peers_connections` method: - For the first 10 attempts, it uses a 1-second sleep between retries. - After 10 attempts, it switches to a 2-second sleep to allow more time for peers to process connections. - Added logging to display the connection status of each peer, making it easier to debug connection issues.

The updated logic now verifies that all nodes are at the same block height and includes a specific check to confirm that Floresta has completed its Initial Block Download (IBD) process.

…n node startup - Added a check to ensure that a node's RPC interface is ready to receive requests after startup. - Specifically for Floresta nodes, added a check to verify that the Electrum server is ready to handle requests when starting for the first time.

Reordered the call to `maybe_open_connection_with_added_peers` to ensure it is the first operation performed during connection setup. This change ensures that the node attempts to connect to manually added peers before performing other validations or attempting connections with other peers.

moisesPompilio · 2026-02-06T15:45:23Z

e4aa3d3 I added the suggestions from the reviews.

Davidson-Souza

ACK e4aa3d3

Phew, that was plenty of work! Good job @moisesPompilio! Now let's go to #742 🫡

jaoleal · 2026-02-09T08:16:32Z

ACK e4aa3d3

But i couldn't really attest that the changes presented here does improve on stability, I had to make a few workarounds to make it work on my end, who happens to be very slow.

Apparently the problem is regarding sleeps around the code where they dont match the required time for the operating nodes to do certain task and when ending the sleep the nodes arent synced. Im trying to come with a fix for that but just some sketchs. (Also not sure if the fix are present in another parallel pr so ill watch after it too.)

Overall, the changes makes sense and improves the framework.

moisesPompilio · 2026-02-09T13:35:48Z

Apparently the problem is regarding sleeps around the code where they dont match the required time for the operating nodes to do certain task and when ending the sleep the nodes arent synced. Im trying to come with a fix for that but just some sketchs. (Also not sure if the fix are present in another parallel pr so ill watch after it too.)

Is the issue caused by sleeps in general, or specifically by sleeps tied to blockchain synchronization (for example, during block mining)?

moisesPompilio mentioned this pull request Dec 16, 2025

Enhance Functional Testing: Daemon Startup Validation, Log Improvements, and Test Restart Refactoring #734

Merged

Davidson-Souza added code quality Generally improves code readability and maintainability functional tests labels Dec 16, 2025

moisesPompilio force-pushed the refactor-functional-architecture branch from 73dd553 to f7fe352 Compare December 29, 2025 17:08

Davidson-Souza reviewed Dec 29, 2025

View reviewed changes

moisesPompilio force-pushed the refactor-functional-architecture branch 2 times, most recently from af27ece to 3440b46 Compare December 30, 2025 19:08

moisesPompilio force-pushed the refactor-functional-architecture branch from 3440b46 to eb485f9 Compare January 7, 2026 20:40

moisesPompilio force-pushed the refactor-functional-architecture branch 3 times, most recently from 53d5ff3 to d3301f7 Compare January 8, 2026 21:53

moisesPompilio force-pushed the refactor-functional-architecture branch from d3301f7 to 0102b58 Compare January 8, 2026 22:12

moisesPompilio mentioned this pull request Jan 12, 2026

Add P2P Communication Validation Tests and Refactor addnode Tests #767

Merged

moisesPompilio force-pushed the refactor-functional-architecture branch from 0102b58 to 4fdba69 Compare January 19, 2026 17:36

moisesPompilio mentioned this pull request Jan 20, 2026

[Enhancement] Improve Integration Tests Coverage and Remove Unnecessary Test Overhead #798

Open

7 tasks

moisesPompilio force-pushed the refactor-functional-architecture branch from 4fdba69 to 240c7ab Compare January 21, 2026 16:08

Davidson-Souza reviewed Jan 21, 2026

View reviewed changes

moisesPompilio force-pushed the refactor-functional-architecture branch 2 times, most recently from ba517ae to ab69cdb Compare January 21, 2026 21:25

Davidson-Souza reviewed Jan 21, 2026

View reviewed changes

moisesPompilio force-pushed the refactor-functional-architecture branch 3 times, most recently from d9f880c to f8c5862 Compare January 26, 2026 21:38

luisschwab assigned moisesPompilio Jan 28, 2026

moisesPompilio force-pushed the refactor-functional-architecture branch 2 times, most recently from e6a0729 to 7a6ac6e Compare February 4, 2026 16:09

moisesPompilio mentioned this pull request Feb 5, 2026

Add Wallet Integration Tests and Fix Descriptor Persistence Issues #821

Open

Davidson-Souza reviewed Feb 5, 2026

View reviewed changes

moisesPompilio added 5 commits February 6, 2026 11:54

refactor(functional): move common RPC methods to BaseRPC

10e51b8

Moved RPC methods that are common across `UtreexoRPC`, `FlorestaRPC`, and `BitcoinRPC` to the `BaseRPC` class to reduce code duplication and centralize shared functionality.

chore(functional): remove invalid Electrum TLS test

0f87eda

Removed `TestSslFailInitialization` from `florestad/tls-fail.py` as it does not test Floresta's behavior. The test only validates `ElectrumClient` connection handling, which is unrelated to Floresta's functionality.

chore(functional): integrate ElectrumClient into Node

b9647aa

The `Node` class now initializes an `ElectrumClient` instance internally, simplifying its usage and removing the need to instantiate it externally.

moisesPompilio force-pushed the refactor-functional-architecture branch from 7a6ac6e to 1eeb017 Compare February 6, 2026 14:54

moisesPompilio added 4 commits February 6, 2026 11:59

chore(integration): refine node synchronization logic in gettxout.py

c7c49bb

The updated logic now verifies that all nodes are at the same block height and includes a specific check to confirm that Floresta has completed its Initial Block Download (IBD) process.

moisesPompilio force-pushed the refactor-functional-architecture branch from 1eeb017 to e4aa3d3 Compare February 6, 2026 15:00

Davidson-Souza approved these changes Feb 6, 2026

View reviewed changes

Davidson-Souza requested a review from jaoleal February 6, 2026 17:31

Davidson-Souza merged commit 07b8f00 into getfloresta:master Feb 9, 2026
11 checks passed

Refactor Test Framework Architecture: Responsibility Segregation, Configuration, and Clarity #733

Refactor Test Framework Architecture: Responsibility Segregation, Configuration, and Clarity #733

Uh oh!

Conversation

moisesPompilio commented Dec 16, 2025

Description and Notes

How to verify the changes you have done?

Uh oh!

Davidson-Souza commented Dec 29, 2025

Uh oh!

luisschwab commented Dec 29, 2025

Uh oh!

moisesPompilio commented Dec 29, 2025

Uh oh!

moisesPompilio commented Dec 29, 2025

Uh oh!

Davidson-Souza left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Davidson-Souza commented Jan 7, 2026

Uh oh!

moisesPompilio commented Jan 7, 2026

Uh oh!

moisesPompilio commented Jan 8, 2026

Uh oh!

Davidson-Souza commented Jan 21, 2026

Uh oh!

moisesPompilio commented Jan 21, 2026

Uh oh!

Davidson-Souza left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

moisesPompilio commented Jan 26, 2026

Uh oh!

moisesPompilio commented Jan 26, 2026

Uh oh!

Davidson-Souza commented Jan 28, 2026

Uh oh!

moisesPompilio commented Feb 4, 2026

Uh oh!

Davidson-Souza commented Feb 4, 2026

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

jaoleal commented Feb 9, 2026 •

edited

Loading