DAOS-6903 Test: Fix regression in NVMe health test code. (#4832)

src/tests/ftest/nvme/enospace.py

@@ -211,6 +211,7 @@ def test_enospace_lazy_with_bg(self):
         #Run IOR to fill the pool.
         self.run_enospace_with_bg_job()
 
+    @skipForTicket("DAOS-7018")
     def test_enospace_lazy_with_fg(self):
         """Jira ID: DAOS-4756.
 
@@ -270,6 +271,7 @@ def test_enospace_time_with_bg(self):
         #Run IOR to fill the pool.
         self.run_enospace_with_bg_job()
 
+    @skipForTicket("DAOS-7018")
     def test_enospace_time_with_fg(self):
         """Jira ID: DAOS-4756.
 
@@ -347,6 +349,7 @@ def test_performance_storage_full(self):
                       ' Baseline Read MiB = {} and latest IOR Read MiB = {}'
                       .format(max_mib_baseline, max_mib_latest))
 
+    @skipForTicket("DAOS-7018")
     def test_enospace_no_aggregation(self):
         """Jira ID: DAOS-4756.
 

src/tests/ftest/nvme/nvme_health.py

@@ -4,22 +4,20 @@
 
   SPDX-License-Identifier: BSD-2-Clause-Patent
 '''
+from __future__ import division
 import os
+
 from nvme_utils import ServerFillUp, get_device_ids
 from test_utils_pool import TestPool
 from dmg_utils import DmgCommand
 from command_utils_base import CommandFailure
-from apricot import skipForTicket
-
 
 class NvmeHealth(ServerFillUp):
     # pylint: disable=too-many-ancestors
     """
     Test Class Description: To validate NVMe health test cases
     :avocado: recursive
     """
-
-    @skipForTicket("DAOS-7011")
     def test_monitor_for_large_pools(self):
         """Jira ID: DAOS-4722.
 
@@ -34,21 +32,21 @@ def test_monitor_for_large_pools(self):
         # pylint: disable=attribute-defined-outside-init
         # pylint: disable=too-many-branches
         no_of_pools = self.params.get("number_of_pools", '/run/pool/*')
-        # Stop the servers to run SPDK too to get the server capacity
-        self.stop_servers()
-        storage = self.get_nvme_max_capacity()
-        self.start_servers()
+        pool_capacity = self.params.get("pool_used_percentage", '/run/pool/*')
+        pool_capacity = pool_capacity / 100
+        storage = self.get_max_storage_sizes()
 
-        # Create the pool from 80% of available of storage space
-        single_pool_nvme_size = int((storage * 0.80)/no_of_pools)
+        #Create the pool from available of storage space
+        single_pool_nvme_size = int((storage[1] * pool_capacity)/no_of_pools)
+        single_pool_scm_size = int((storage[0] * pool_capacity)/no_of_pools)
 
         self.pool = []
         # Create the Large number of pools
         for _pool in range(no_of_pools):
+            self.log.info("-- Creating pool number = %s", _pool)
             pool = TestPool(self.context, self.get_dmg_command())
             pool.get_params(self)
-            # SCM size is 10% of NVMe
-            pool.scm_size.update('{}'.format(int(single_pool_nvme_size * 0.10)))
+            pool.scm_size.update('{}'.format(single_pool_scm_size))
             pool.nvme_size.update('{}'.format(single_pool_nvme_size))
             pool.create()
             self.pool.append(pool)

src/tests/ftest/nvme/nvme_health.yaml

@@ -44,3 +44,4 @@ pool:
     name: daos_server
     control_method: dmg
     number_of_pools: 40
+    pool_used_percentage: 75

src/tests/ftest/nvme/nvme_io.yaml

@@ -28,6 +28,7 @@ pool:
 ior:
  flags: "-w -r -k -vv"
  repetitions: 1
+ test_file: /testFile
  object_type:
   - "SX"
   - "S1"
@@ -41,7 +42,7 @@ ior:
  dfs_destroy: False
  ior_sequence:
 #   - [scmsize, nvmesize, stripesize, blocksize, clientslots]
-    - [6000000000, 139586437120, 1048576, 17179869184, 1]        #[6G, 130G, 1M, 16G, 1]
+    - [6000000000, 214748364800, 1048576, 17179869184, 1]        #[6G, 200G, 1M, 16G, 1]
     - [1000000000, 9000000000, 4096, 1073741824, 1]              #[1G, 9G, 4k, 1G, 1]
     - [10000000000, 9000000000, 2048, 1073741824, 1]              #[10G, 9G, 2k, 1G, 1]
     - [3000000000, 9000000000, 8, 10485760, 1]                   #[3G, 9G, 8B, 10M, 1]

src/tests/ftest/util/general_utils.py

@@ -853,7 +853,7 @@ def error_count(error, hostlist, log_file):
     # Get the Client side Error from client_log file.
     requested_error_count = 0
     other_error_count = 0
-    command = 'cat {} | grep ERR'.format(get_log_file(log_file))
+    command = 'cat {} | grep \" ERR \"'.format(get_log_file(log_file))
     results = run_pcmd(hostlist, command, False, None, None)
     for result in results:
         for line in result["stdout"]:
@@ -865,7 +865,6 @@ def error_count(error, hostlist, log_file):
 
     return requested_error_count, other_error_count
 
-
 def get_module_class(name, module):
     """Get the class object in the specified module by its name.