DAOS-13882 test: Continue archiving on successful hosts. (#13110)

If an error is encountered on some of the hosts during the archiving of
the test results, continue the remaining archiving steps on the
successful hosts.

Reorganize a few launch.py methods for future maintenance.

Launch.py now handles the xml results classname renaming for the cmocka
files generated by the tests.  This includes referencing the python test
class used to run the test in the cmocka results classname.

Also include cleaning up multiple testsuite cmocka xml entries.

List tests in PR that don't match tags.

Fix utility imports.
Add ftest to PYTHONPATH for import util.<file> to work.

Signed-off-by: Phil Henderson <[email protected]>

src/tests/ftest/control/dmg_network_scan.py

@@ -4,7 +4,9 @@
   SPDX-License-Identifier: BSD-2-Clause-Patent
 """
 from apricot import TestWithServers
-from network_utils import SUPPORTED_PROVIDERS, get_dmg_network_information, get_network_information
+from ClusterShell.NodeSet import NodeSet
+from exception_utils import CommandFailure
+from network_utils import SUPPORTED_PROVIDERS, NetworkDevice, get_network_information
 
 
 class DmgNetworkScanTest(TestWithServers):
@@ -29,7 +31,7 @@ def get_sys_info(self):
         """
         server_provider = self.server_managers[0].get_config_value("provider")
         sys_info = []
-        for entry in get_network_information(self.hostlist_servers, SUPPORTED_PROVIDERS):
+        for entry in get_network_information(self.log, self.hostlist_servers, SUPPORTED_PROVIDERS):
             if server_provider in entry.provider:
                 entry.device = None
                 sys_info.append(entry)
@@ -43,7 +45,37 @@ def get_dmg_info(self):
 
         """
         dmg = self.get_dmg_command()
-        return get_dmg_network_information(dmg.network_scan())
+        return self.get_dmg_network_information(dmg.network_scan())
+
+    def get_dmg_network_information(self, dmg_network_scan):
+        """Get the network device information from the dmg network scan output.
+
+        Args:
+            dmg_network_scan (dict): the dmg network scan json command output
+
+        Raises:
+            CommandFailure: if there was an error processing the dmg network scan output
+
+        Returns:
+            list: a list of NetworkDevice objects identifying the network devices on each host
+
+        """
+        network_devices = []
+
+        try:
+            for host_fabric in dmg_network_scan["response"]["HostFabrics"].values():
+                for host in NodeSet(host_fabric["HostSet"].split(":")[0]):
+                    for interface in host_fabric["HostFabric"]["Interfaces"]:
+                        network_devices.append(
+                            NetworkDevice(
+                                host, interface["Device"], None, 1, interface["Provider"],
+                                interface["NumaNode"])
+                        )
+        except KeyError as error:
+            raise CommandFailure(
+                f"Error processing dmg network scan json output: {dmg_network_scan}") from error
+
+        return network_devices
 
     def test_dmg_network_scan_basic(self):
         """JIRA ID: DAOS-2516

src/tests/ftest/deployment/network_failure.py

@@ -38,7 +38,7 @@ def pre_tear_down(self):
             self.log.debug("Call ip link set before tearDown.")
             if self.network_down_host:
                 update_network_interface(
-                    interface=self.interface, state="up", hosts=self.network_down_host,
+                    self.log, interface=self.interface, state="up", hosts=self.network_down_host,
                     errors=error_list)
         return error_list
 
@@ -206,7 +206,7 @@ def verify_network_failure(self, ior_namespace, container_namespace):
         if self.test_env == "ci":
             # wolf
             update_network_interface(
-                interface=self.interface, state="down", hosts=self.network_down_host,
+                self.log, interface=self.interface, state="down", hosts=self.network_down_host,
                 errors=errors)
         else:
             # Aurora. Manually run the command.
@@ -228,7 +228,7 @@ def verify_network_failure(self, ior_namespace, container_namespace):
         if self.test_env == "ci":
             # wolf
             update_network_interface(
-                interface=self.interface, state="up", hosts=self.network_down_host,
+                self.log, interface=self.interface, state="up", hosts=self.network_down_host,
                 errors=errors)
         else:
             # Aurora. Manually run the command.
@@ -392,7 +392,7 @@ def test_network_failure_isolation(self):
         # wolf
         if self.test_env == "ci":
             update_network_interface(
-                interface=self.interface, state="down", hosts=self.network_down_host,
+                self.log, interface=self.interface, state="down", hosts=self.network_down_host,
                 errors=errors)
         else:
             # Aurora. Manually run the command.
@@ -431,7 +431,7 @@ def test_network_failure_isolation(self):
         if self.test_env == "ci":
             # wolf
             update_network_interface(
-                interface=self.interface, state="up", hosts=self.network_down_host,
+                self.log, interface=self.interface, state="up", hosts=self.network_down_host,
                 errors=errors)
         else:
             # Aurora. Manually run the command.

src/tests/ftest/harness/core_files.py

@@ -52,7 +52,7 @@ def test_core_files(self):
             with open(core_file, "w", encoding="utf-8") as local_core_file:
                 local_core_file.write("THIS IS JUST A TEST\n")
         except IOError as error:
-            self.fail("Error writing {}: {}".format(local_core_file, str(error)))
+            self.fail("Error writing {}: {}".format(core_file, str(error)))
 
         # Choose a server find the pid of its daos_engine process
         host = NodeSet(choice(self.server_managers[0].hosts))   # nosec