diff --git a/letta/services/tool_execution_sandbox.py b/letta/services/tool_execution_sandbox.py
index b05b6ce2..1060af43 100644
--- a/letta/services/tool_execution_sandbox.py
+++ b/letta/services/tool_execution_sandbox.py
@@ -277,6 +277,10 @@ class ToolExecutionSandbox:
         sbx_config = self.sandbox_config_manager.get_or_create_default_sandbox_config(sandbox_type=SandboxType.E2B, actor=self.user)
         sbx = self.get_running_e2b_sandbox_with_same_state(sbx_config)
         if not sbx or self.force_recreate:
+            if not sbx:
+              logger.info(f"No running e2b sandbox found with the same state: {sbx_config}")
+            else:
+              logger.info(f"Force recreated e2b sandbox with state: {sbx_config}")
             sbx = self.create_e2b_sandbox_with_metadata_hash(sandbox_config=sbx_config)
 
         # Since this sandbox was used, we extend its lifecycle by the timeout
@@ -292,6 +296,8 @@ class ToolExecutionSandbox:
             func_return, agent_state = self.parse_best_effort(execution.results[0].text)
         elif execution.error:
             logger.error(f"Executing tool {self.tool_name} failed with {execution.error}")
+            logger.error(f"E2B Sandbox configurations: {sbx_config}")
+            logger.error(f"E2B Sandbox ID: {sbx.sandbox_id}")
             func_return = get_friendly_error_msg(
                 function_name=self.tool_name, exception_name=execution.error.name, exception_message=execution.error.value
             )
diff --git a/tests/integration_test_tool_execution_sandbox.py b/tests/integration_test_tool_execution_sandbox.py
index ad3d2ebb..3f64b287 100644
--- a/tests/integration_test_tool_execution_sandbox.py
+++ b/tests/integration_test_tool_execution_sandbox.py
@@ -464,7 +464,7 @@ def test_e2b_sandbox_inject_env_var_existing_sandbox(check_e2b_key_is_set, get_e
     config = manager.create_or_update_sandbox_config(config_create, test_user)
 
     # Run the custom sandbox once, assert nothing returns because missing env variable
-    sandbox = ToolExecutionSandbox(get_env_tool.name, {}, user=test_user, force_recreate=True)
+    sandbox = ToolExecutionSandbox(get_env_tool.name, {}, user=test_user)
     result = sandbox.run()
     # response should be None
     assert result.func_return is None
diff --git a/tests/test_client.py b/tests/test_client.py
index c3e507ae..532d8058 100644
--- a/tests/test_client.py
+++ b/tests/test_client.py
@@ -42,8 +42,8 @@ def run_server():
 
 
 @pytest.fixture(
-    # params=[{"server": False}, {"server": True}],  # whether to use REST API server
-    params=[{"server": False}],  # whether to use REST API server
+    params=[{"server": False}, {"server": True}],  # whether to use REST API server
+    # params=[{"server": False}],  # whether to use REST API server
     scope="module",
 )
 def client(request):
@@ -385,7 +385,7 @@ def test_function_return_limit(mock_e2b_api_key_none, client: Union[LocalClient,
     client.delete_agent(agent_id=agent.id)
 
 
-def test_function_always_error(mock_e2b_api_key_none, client: Union[LocalClient, RESTClient]):
+def test_function_always_error(client: Union[LocalClient, RESTClient]):
     """Test to see if function that errors works correctly"""
 
     def always_error():
@@ -409,7 +409,6 @@ def test_function_always_error(mock_e2b_api_key_none, client: Union[LocalClient,
     assert response_message, "ToolReturnMessage message not found in response"
     assert response_message.status == "error"
 
-    import ipdb;ipdb.set_trace()
     if isinstance(client, RESTClient):
         assert response_message.tool_return == "Error executing function always_error: ZeroDivisionError: division by zero"
     else: