tests/consensus_tests/test_force_delete_peer.py (5)

11-16: Add docstring for helper function.

This function is well-implemented, but lacks documentation explaining its purpose and behavior.

 def force_delete_peer(peer_api_uri: str, peer_id: int):
+    """
+    Force deletes a peer from the cluster.
+    
+    Args:
+        peer_api_uri: URI of the peer API to send the request to
+        peer_id: ID of the peer to delete
+    
+    Raises:
+        AssertionError: If the request fails
+    """
     response = requests.delete(
         f"{peer_api_uri}/cluster/peer/{peer_id}?force=true",
     )
     assert response.status_code == 200, f"Failed to force delete peer: {response.text}"

🧰 Tools

🪛 Ruff (0.8.2)

12-12: requests may be undefined, or defined from star imports

(F405)

---

18-22: Add docstring for get_peer_id function.

Similar to the previous function, this helper function would benefit from documentation explaining its purpose and return value.

 def get_peer_id(peer_api_uri: str) -> int:
+    """
+    Gets the peer ID from the cluster info endpoint.
+    
+    Args:
+        peer_api_uri: URI of the peer API to send the request to
+        
+    Returns:
+        int: The ID of the peer
+        
+    Raises:
+        AssertionError: If the request fails
+    """
     response = requests.get(f"{peer_api_uri}/cluster")
     assert response.status_code == 200, f"Failed to get peer ID: {response.text}"
     return response.json()["result"]["peer_id"]

🧰 Tools

🪛 Ruff (0.8.2)

19-19: requests may be undefined, or defined from star imports

(F405)

---

42-46: Add comment explaining which peer is being killed.

The comment "Kill last peer" is somewhat ambiguous. Add more context about why this specific peer is chosen and what effect it will have.

-    # Now
-    # Kill last peer
+    # Kill the last peer to simulate a node failure
+    # This will cause its replicas to be marked as dead
     p = processes.pop()
     p.kill()
     sleep(1)  # Give killed peer time to release WAL lock

🧰 Tools

🪛 Ruff (0.8.2)

44-44: processes may be undefined, or defined from star imports

(F405)

---

48-49: Clarify the purpose of additional data insertion.

The comment doesn't fully explain why inserting more points triggers recovery transfer.

-    # Upsert points to mark dummy replica as dead, that will trigger recovery transfer
+    # Upsert points to trigger write operations that will mark the killed peer's replicas
+    # as dead, which will then trigger recovery transfers when the peer is restarted
     upsert_random_points(peer_api_uris[0], 100)

---

78-84: Add assertion error message and consider using wait function.

Consider adding an error message to make test failures more informative, and consider using a wait function instead of a fixed sleep to make the test more reliable.

-    sleep(1)
+    # Wait a moment for the force delete operation to be processed
+    sleep(1)
 
     # We expect transfers to be aborted
     transfers = get_collection_cluster_info(peer_api_uris[0], COLLECTION_NAME)[
         "shard_transfers"
     ]
-    assert len(transfers) == 0
+    assert len(transfers) == 0, "Shard transfers were not aborted after force deleting the source peer"

Consider using a wait function for more reliable testing:

def wait_for_no_transfers(peer_api_uri: str, collection_name: str, timeout: int = 10):
    """Wait until there are no shard transfers for the collection."""
    start_time = time.time()
    while time.time() - start_time < timeout:
        transfers = get_collection_cluster_info(peer_api_uri, collection_name)["shard_transfers"]
        if len(transfers) == 0:
            return
        sleep(0.5)
    raise TimeoutError(f"Transfers were not aborted within {timeout} seconds")

# Then use it instead of sleep + assert
wait_for_no_transfers(peer_api_uris[0], COLLECTION_NAME)

🧰 Tools

🪛 Ruff (0.8.2)

81-81: get_collection_cluster_info may be undefined, or defined from star imports

(F405)

Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro

tests/consensus_tests/test_force_delete_peer.py (3)

2-2: Avoid wildcard imports.

The wildcard import (from .utils import *) makes it difficult to track which symbols are being imported and where they come from. This can lead to naming conflicts and makes the code harder to understand and maintain.

-from .utils import *
+from .utils import (assert_project_root, start_cluster, start_peer, 
+                    wait_collection_exists_and_active_on_all_peers, 
+                    wait_for_collection_shard_transfers_count, 
+                    get_collection_cluster_info, processes)

🧰 Tools

🪛 Ruff (0.8.2)

2-2: from .utils import * used; unable to detect undefined names

(F403)

---

23-23: Add docstring to explain test purpose and steps.

The test function should include a docstring that explains its purpose and the steps it follows to test the behavior.

def test_force_delete_source_peer_during_transfers(tmp_path: pathlib.Path):
+    """
+    Tests that shard transfers are aborted when the source peer is force deleted.
+    
+    Test steps:
+    1. Start a cluster with multiple peers
+    2. Create a collection with multiple shards and replicas
+    3. Insert initial data
+    4. Kill one peer to simulate failure
+    5. Insert more data to trigger recovery transfer
+    6. Restart the killed peer
+    7. Wait for shard transfer to start
+    8. Force delete the source peer of the transfer
+    9. Verify that transfers are aborted
+    """
    assert_project_root()

---

67-72: Handle the unused variable and clarify comment.

The variable url is assigned but never used, and the comment could be clearer about the purpose of stopping the source node.

-    # Stop the 'source' node to simulate an unreachable node
+    # Stop the source node of the shard transfer to simulate a failure during transfer
     source_peer_url = peer_id_to_url[from_peer_id]
     peer_idx = peer_api_uris.index(source_peer_url)
     p = processes.pop(peer_idx)
-    url = peer_api_uris.pop(peer_idx)
+    peer_api_uris.pop(peer_idx)  # Remove the URL of the killed peer from the list
     sleep(1)  # Give killed peer time to release WAL lock

🧰 Tools

🪛 Ruff (0.8.2)

70-70: processes may be undefined, or defined from star imports

(F405)

---

71-71: Local variable url is assigned to but never used

Remove assignment to unused variable url

(F841)

Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro

tests/consensus_tests/test_force_delete_peer.py (3)

1-4: Replace wildcard imports with explicit imports for better maintainability.

Using wildcard imports (from .utils import *) makes it difficult to track which symbols are being imported. This can lead to naming conflicts and makes the code harder to understand and maintain.

 import pathlib
-from .utils import *
+from .utils import (
+    assert_project_root, start_cluster, start_peer, processes,
+    wait_collection_exists_and_active_on_all_peers,
+    wait_for_collection_shard_transfers_count,
+    get_collection_cluster_info
+)
 from .fixtures import upsert_random_points, create_collection
 from time import sleep
+import requests

🧰 Tools

🪛 Ruff (0.8.2)

2-2: from .utils import * used; unable to detect undefined names

(F403)

---

23-23: Add docstring to explain test purpose and steps.

Adding a docstring would make the test function's purpose and methodology clearer to other developers.

 def test_force_delete_source_peer_during_transfers(tmp_path: pathlib.Path):
+    """
+    Tests that shard transfers are aborted when the source peer is force deleted.
+    
+    Test steps:
+    1. Start a cluster with multiple peers
+    2. Create a collection with multiple shards and replicas
+    3. Insert initial data
+    4. Kill one peer to simulate failure
+    5. Insert more data to trigger recovery transfer
+    6. Restart the killed peer
+    7. Wait for shard transfer to start
+    8. Force delete the source peer of the transfer
+    9. Verify that transfers are aborted
+    """
     assert_project_root()

---

66-70: Fix unused variable and clarify comment about stopping the source node.

The variable url is assigned but never used. Also, the comment could be more specific about why you're stopping the source node.

-    # Stop the 'source' node to simulate an unreachable node
+    # Stop the source node of the shard transfer to simulate a failure during transfer
     source_peer_url = peer_id_to_url[from_peer_id]
     peer_idx = peer_api_uris.index(source_peer_url)
     p = processes.pop(peer_idx)
-    url = peer_api_uris.pop(peer_idx)
+    peer_api_uris.pop(peer_idx)  # Remove the URL of the killed peer from the list

🧰 Tools

🪛 Ruff (0.8.2)

69-69: processes may be undefined, or defined from star imports

(F405)

---

70-70: Local variable url is assigned to but never used

Remove assignment to unused variable url

(F841)

tests/consensus_tests/test_force_delete_peer.py (3)

63-63: Improve assertion message with more context.

Add an assertion message to make it clearer what's being tested and why the last peer is expected to be the destination.

-    assert transfers[0]["to"] == list(peer_id_to_url.keys())[-1] # last peer was restarted
+    assert transfers[0]["to"] == list(peer_id_to_url.keys())[-1], "Transfer should be to the restarted peer"

---

75-75: Replace arbitrary sleep with a more reliable wait mechanism.

Using a fixed sleep duration can make tests flaky. Consider implementing a polling mechanism that waits for a specific condition or explain why the sleep is necessary in this case.

-    sleep(1)
+    # Wait briefly for cluster state to update after peer deletion
+    sleep(1)  # This delay is necessary to ensure cluster state is updated before checking transfers

---

77-80: Add a detailed assertion message for test conclusion.

Make the final assertion more informative by explaining the expected behavior.

     # We expect transfers to be aborted
     wait_for_collection_shard_transfers_count(
-        peer_api_uris[0], COLLECTION_NAME, 0
+        peer_api_uris[0], COLLECTION_NAME, 0,
+        # Verification that after force-deleting the source peer, no shard transfers remain
     )

🧰 Tools

🪛 Ruff (0.8.2)

78-78: wait_for_collection_shard_transfers_count may be undefined, or defined from star imports

(F405)

Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro

tests/consensus_tests/test_force_delete_peer.py (1)

67-71: 🛠️ Refactor suggestion

Add a timeout to the transfer abort verification.

If the transfers are not aborted as expected, the current wait function might hang indefinitely. Consider adding a timeout parameter or checking for specific failure conditions.

     # We expect transfers to be aborted
     wait_for_collection_shard_transfers_count(
-        peer_api_uris[0], COLLECTION_NAME, 0
+        peer_api_uris[0], COLLECTION_NAME, 0, max_wait=30  # Add a timeout of 30 seconds
     )
+    
+    # Verify the peer was actually deleted
+    cluster_info = requests.get(f"{peer_api_uris[0]}/cluster").json()["result"]
+    assert from_peer_id not in [peer["id"] for peer in cluster_info["peers"]], "Source peer was not actually deleted"

🧰 Tools

🪛 Ruff (0.8.2)

68-68: wait_for_collection_shard_transfers_count may be undefined, or defined from star imports

(F405)

tests/consensus_tests/test_force_delete_peer.py (2)

1-5: Add explicit imports for clarity and maintainability.

The code uses wildcard imports and implicitly relies on the requests module that's imported through the star import. For better code clarity, maintainability, and to avoid potential issues if the utility module changes, use explicit imports.

import pathlib
from .utils import *
from .fixtures import upsert_random_points, create_collection
from time import sleep
+import requests

Additionally, I recommend explicitly importing the specific functions you're using from .utils rather than using a wildcard import:

-from .utils import *
+from .utils import (
+    assert_project_root, start_cluster, start_peer,
+    wait_collection_exists_and_active_on_all_peers,
+    wait_for_collection_shard_transfers_count,
+    get_collection_cluster_info, processes
+)

🧰 Tools

🪛 Ruff (0.8.2)

2-2: from .utils import * used; unable to detect undefined names

(F403)

---

21-29: Add a docstring to explain test purpose and steps.

The test function should include a docstring that explains its purpose and the steps it follows to test the behavior.

def test_force_delete_source_peer_during_transfers(tmp_path: pathlib.Path):
+    """
+    Tests that shard transfers are aborted when the source peer is force deleted.
+    
+    Test steps:
+    1. Start a cluster with multiple peers
+    2. Create a collection with multiple shards and replicas
+    3. Insert initial data
+    4. Kill one peer to simulate failure
+    5. Insert more data to trigger recovery transfer
+    6. Restart the killed peer
+    7. Wait for shard transfer to start
+    8. Force delete the source peer of the transfer
+    9. Verify that transfers are aborted
+    """
    assert_project_root()

    peer_api_uris, peer_dirs, bootstrap_uri = start_cluster(tmp_path, N_PEERS)

🧰 Tools

🪛 Ruff (0.8.2)

22-22: assert_project_root may be undefined, or defined from star imports

(F405)

---

24-24: start_cluster may be undefined, or defined from star imports

(F405)

---

27-27: wait_collection_exists_and_active_on_all_peers may be undefined, or defined from star imports

(F405)

tests/consensus_tests/test_force_delete_peer.py (4)

9-14: Function looks good but add a docstring.

The helper function is well-implemented with proper error handling. Consider adding a docstring to explain the purpose of this function and its parameters.

def force_delete_peer(peer_api_uri: str, peer_id: int):
+    """
+    Force delete a peer from the cluster.
+    
+    Args:
+        peer_api_uri: The API URI of any active peer in the cluster
+        peer_id: The ID of the peer to be deleted
+    
+    Raises:
+        AssertionError: If the request fails
+    """
    response = requests.delete(
        f"{peer_api_uri}/cluster/peer/{peer_id}?force=true",
    )
    assert response.status_code == 200, f"Failed to force delete peer: {response.text}"

🧰 Tools

🪛 Ruff (0.8.2)

10-10: requests may be undefined, or defined from star imports

(F405)

---

16-20: Add a docstring to the helper function.

Similar to the previous function, adding a docstring would improve code readability and maintainability.

def get_peer_id(peer_api_uri: str) -> int:
+    """
+    Get the ID of a peer by querying its cluster information.
+    
+    Args:
+        peer_api_uri: The API URI of the peer
+        
+    Returns:
+        int: The ID of the peer
+        
+    Raises:
+        AssertionError: If the request fails
+    """
    response = requests.get(f"{peer_api_uri}/cluster")
    assert response.status_code == 200, f"Failed to get peer ID: {response.text}"
    return response.json()["result"]["peer_id"]

🧰 Tools

🪛 Ruff (0.8.2)

17-17: requests may be undefined, or defined from star imports

(F405)

---

39-43: Clarify the peer selection strategy.

The test kills the last peer from the processes list but there's no explicit verification that this corresponds to the last peer in peer_api_uris. A comment would help clarify this relationship.

-    # Kill last peer
+    # Kill last peer (processes and peer_api_uris are populated in the same order by start_cluster)
     p = processes.pop()
     p.kill()
     sleep(1)  # Give killed peer time to release WAL lock

🧰 Tools

🪛 Ruff (0.8.2)

40-40: processes may be undefined, or defined from star imports

(F405)

---

54-61: Add verification of the shard transfer.

The test waits for a shard transfer to start and checks the transfer count, but it would be good to add more detailed verification of the transfer state and logging for debugging.

     # Wait for start of shard transfer
     wait_for_collection_shard_transfers_count(peer_api_uris[0], COLLECTION_NAME, 1)

     transfers = get_collection_cluster_info(peer_api_uris[0], COLLECTION_NAME)[
         "shard_transfers"
     ]
     assert len(transfers) == 1
+    print(f"Observed shard transfer: {transfers[0]}")
     assert transfers[0]["to"] == list(peer_id_to_url.keys())[-1] # last peer was restarted
+    assert "from" in transfers[0], f"Missing source peer in transfer: {transfers[0]}"
     from_peer_id = transfers[0]["from"]

🧰 Tools

🪛 Ruff (0.8.2)

55-55: wait_for_collection_shard_transfers_count may be undefined, or defined from star imports

(F405)

---

57-57: get_collection_cluster_info may be undefined, or defined from star imports

(F405)

Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro

tests/consensus_tests/test_force_delete_peer.py (3)

1-5: Explicit imports should be used instead of wildcard imports.

The code currently uses a wildcard import (from .utils import *) which makes it difficult to track which symbols are being imported. This could lead to maintainability issues and potential naming conflicts. Additionally, the code uses the requests module without explicitly importing it.

import pathlib
-from .utils import *
+from .utils import (assert_project_root, start_cluster, start_peer, 
+                    wait_collection_exists_and_active_on_all_peers, 
+                    wait_for_collection_shard_transfers_count, 
+                    get_collection_cluster_info, wait_for_same_commit,
+                    processes)
from .fixtures import upsert_random_points, create_collection
from time import sleep
+import requests

🧰 Tools

🪛 Ruff (0.8.2)

2-2: from .utils import * used; unable to detect undefined names

(F403)

---

21-22: Add a docstring to the test function.

A docstring would help explain the purpose of the test and the steps it takes, making it easier for other developers to understand the test case.

def test_force_delete_source_peer_during_transfers(tmp_path: pathlib.Path):
+    """
+    Tests that shard transfers are aborted when the source peer is force deleted.
+    
+    Test steps:
+    1. Start a cluster with multiple peers
+    2. Create a collection with multiple shards and replicas
+    3. Insert initial data
+    4. Kill one peer to simulate failure
+    5. Insert more data to trigger recovery transfer
+    6. Restart the killed peer
+    7. Wait for shard transfer to start
+    8. Force delete the source peer of the transfer
+    9. Verify that transfers are aborted
+    """
    assert_project_root()

🧰 Tools

🪛 Ruff (0.8.2)

22-22: assert_project_root may be undefined, or defined from star imports

(F405)

---

64-69: Improve the comment and handling of the source peer removal.

The comment could be clearer and the code for killing the peer could be more explicit.

-    # Stop the 'source' node to simulate an unreachable node which needs to be deleted
+    # Stop the source node of the shard transfer to simulate a failure during transfer
     source_peer_url = peer_id_to_url[from_peer_id]
     peer_idx = peer_api_uris.index(source_peer_url)
-    processes.pop(peer_idx).kill() # Kill the source peer
-    peer_api_uris.pop(peer_idx) # Remove from urls so we don't try to call it
+    source_process = processes.pop(peer_idx) 
+    source_process.kill()  # Kill the source peer
+    peer_api_uris.pop(peer_idx)  # Remove from urls so we don't try to call it

🧰 Tools

🪛 Ruff (0.8.2)

67-67: processes may be undefined, or defined from star imports

(F405)

tests/consensus_tests/test_force_delete_peer.py (3)

39-42: Add a more robust way to identify the peer to kill.

The current implementation simply kills the last peer in the processes list, which might be brittle if the order of processes changes. Consider using a more explicit way to identify which peer to kill.

-    # Kill last peer
-    p = processes.pop()
-    p.kill()
+    # Kill the last peer to simulate a node failure
+    last_peer_process = processes.pop()
+    last_peer_process.kill()
     sleep(1)  # Give killed peer time to release WAL lock

🧰 Tools

🪛 Ruff (0.8.2)

40-40: processes may be undefined, or defined from star imports

(F405)

---

57-63: Verify the data transfer properties more thoroughly.

The assertions are good, but could be more comprehensive to ensure the transfer is happening as expected. Consider adding assertions for other properties of the transfer.

    transfers = get_collection_cluster_info(peer_api_uris[0], COLLECTION_NAME)[
        "shard_transfers"
    ]
    assert len(transfers) == 1
+    # Verify that transfer is going to the restarted peer
    assert transfers[0]["to"] == list(peer_id_to_url.keys())[-1] # last peer was restarted
+    # Verify that transfer is happening from a valid source peer
    from_peer_id = transfers[0]["from"]
+    assert from_peer_id in peer_id_to_url, f"Invalid source peer ID: {from_peer_id}"

🧰 Tools

🪛 Ruff (0.8.2)

57-57: get_collection_cluster_info may be undefined, or defined from star imports

(F405)

---

75-78: Add an assertion to verify cluster stability after force deleting a peer.

After verifying that transfers are aborted, it would be valuable to ensure the cluster remains stable and operational.

    # We expect transfers to be aborted
    wait_for_collection_shard_transfers_count(
        peer_api_uris[0], COLLECTION_NAME, 0
    )
+    
+    # Verify that the cluster remains stable after force deleting a peer
+    # Try to insert some more points to ensure the cluster is still operational
+    upsert_random_points(peer_api_uris[0], 50)
+    
+    # Verify that all remaining peers have the same view of the cluster
+    wait_for_same_commit(peer_api_uris)

🧰 Tools

🪛 Ruff (0.8.2)

76-76: wait_for_collection_shard_transfers_count may be undefined, or defined from star imports

(F405)

Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro

tests/consensus_tests/test_force_delete_peer.py (2)

1-5: 🛠️ Refactor suggestion

Clean up imports to improve code maintainability.

The code uses wildcard imports (from .utils import *) which makes it difficult to track which symbols are imported. It also uses the requests module without explicitly importing it.

import pathlib
-from .utils import *
+from .utils import (assert_project_root, start_cluster, start_peer, 
+                    wait_collection_exists_and_active_on_all_peers, 
+                    wait_for_collection_shard_transfers_count, 
+                    get_collection_cluster_info, processes)
from .fixtures import upsert_random_points, create_collection
from time import sleep
+import requests

🧰 Tools

🪛 Ruff (0.8.2)

2-2: from .utils import * used; unable to detect undefined names

(F403)

---

21-22: 🛠️ Refactor suggestion

Add docstring to explain test purpose and steps.

The test function should include a docstring that explains its purpose and the steps it follows to test the behavior.

def test_force_delete_source_peer_during_transfers(tmp_path: pathlib.Path):
+    """
+    Tests that shard transfers are aborted when the source peer is force deleted.
+    
+    Test steps:
+    1. Start a cluster with 3 peers
+    2. Create a collection with 2 shards and 3 replicas
+    3. Insert initial data
+    4. Kill one peer to simulate failure
+    5. Insert more data to trigger recovery transfer
+    6. Restart the killed peer
+    7. Wait for shard transfer to start
+    8. Force delete the source peer of the transfer
+    9. Verify that transfers are aborted
+    """
    assert_project_root()

🧰 Tools

🪛 Ruff (0.8.2)

22-22: assert_project_root may be undefined, or defined from star imports

(F405)

tests/consensus_tests/test_force_delete_peer.py (5)

9-14: Function looks good, but add docstring.

The function to force delete a peer is well-implemented, but missing a docstring to explain its purpose and parameters.

def force_delete_peer(peer_api_uri: str, peer_id: int):
+    """
+    Force delete a peer from the cluster with the specified ID.
+    
+    Args:
+        peer_api_uri: URI of any active peer in the cluster
+        peer_id: ID of the peer to be forcibly deleted
+    """
    response = requests.delete(
        f"{peer_api_uri}/cluster/peer/{peer_id}?force=true",
    )
    assert response.status_code == 200, f"Failed to force delete peer: {response.text}"

🧰 Tools

🪛 Ruff (0.8.2)

10-10: requests may be undefined, or defined from star imports

(F405)

---

16-20: Function looks good, but add docstring.

The function to get peer ID is well-implemented, but missing a docstring to explain its purpose and parameters.

def get_peer_id(peer_api_uri: str) -> int:
+    """
+    Get the peer ID for a peer given its API URI.
+    
+    Args:
+        peer_api_uri: URI of the peer
+        
+    Returns:
+        The peer ID as an integer
+    """
    response = requests.get(f"{peer_api_uri}/cluster")
    assert response.status_code == 200, f"Failed to get peer ID: {response.text}"
    return response.json()["result"]["peer_id"]

🧰 Tools

🪛 Ruff (0.8.2)

17-17: requests may be undefined, or defined from star imports

(F405)

---

39-42: Improve comment clarity for killed peer.

The comment "Kill last peer" could be clearer about which peer is being killed and why.

-    # Kill last peer
+    # Kill the last peer to simulate a node failure
     p = processes.pop()
     p.kill()
     sleep(1)  # Give killed peer time to release WAL lock

🧰 Tools

🪛 Ruff (0.8.2)

40-40: processes may be undefined, or defined from star imports

(F405)

---

44-45: Clarify why upserting points triggers recovery.

The comment mentions marking a dummy replica as dead, but it would be helpful to explain how upserting points accomplishes this.

-    # Upsert points to mark dummy replica as dead, that will trigger recovery transfer
+    # Upsert points - since one peer is down, these updates won't reach it
+    # This will mark its replicas as stale, triggering recovery transfer when it rejoins
     upsert_random_points(peer_api_uris[0], 100)

---

54-62: Good verification of transfer target, consider verifying shard ID.

The code correctly waits for and verifies the shard transfer, but it might be useful to capture and verify the specific shard ID being transferred.

    # Wait for start of shard transfer
    wait_for_collection_shard_transfers_count(peer_api_uris[0], COLLECTION_NAME, 1)

    transfers = get_collection_cluster_info(peer_api_uris[0], COLLECTION_NAME)[
        "shard_transfers"
    ]
    assert len(transfers) == 1
    assert transfers[0]["to"] == list(peer_id_to_url.keys())[-1] # last peer was restarted
    from_peer_id = transfers[0]["from"]
+    shard_id = transfers[0]["shard_id"]  # Capture shard ID for debugging/verification

🧰 Tools

🪛 Ruff (0.8.2)

55-55: wait_for_collection_shard_transfers_count may be undefined, or defined from star imports

(F405)

---

57-57: get_collection_cluster_info may be undefined, or defined from star imports

(F405)

Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro