Merge pull request #13 from jerewill-cisco/pre-release

Release for 1.2.7

Author: jerewill-cisco

input_module_intersight.py

@@ -184,10 +184,15 @@ def collect_events(helper, ew):
             helper.log_debug(stanza_name + ' | ' +
                              "Creating event for Moid "+data['Moid'])
             # Here we check to see if the latest event is newer than our checkpoint, if so we update it.
-            if datetime.datetime.strptime(state, "%Y-%m-%dT%H:%M:%S.%f%z") < datetime.datetime.strptime(data['ModTime'], "%Y-%m-%dT%H:%M:%S.%f%z"):
-                state = data['ModTime']
+            try:
+                if datetime.datetime.strptime(state, "%Y-%m-%dT%H:%M:%S.%f%z") < datetime.datetime.strptime(data['ModTime'], "%Y-%m-%dT%H:%M:%S.%f%z"):
+                    state = data['ModTime']
+                    helper.log_debug(
+                        stanza_name + ' | ' + "Checkpoint key for audit records was updated to " + state)
+            except (ValueError):
                 helper.log_debug(
-                    stanza_name + ' | ' + "Checkpoint key for audit records was updated to " + state)
+                        stanza_name + ' | ' + "Checkpoint key for audit records was unable to be updated")
+                pass
         # Persist our checkpoint at the end of the audit records
         helper.save_check_point(account_name+'_last_audit_record', state)
 
@@ -257,10 +262,15 @@ def collect_events(helper, ew):
             helper.log_debug(stanza_name + ' | ' +
                              "Creating event for Moid "+data['Moid'])
             # Here we check to see if the latest event is newer than our checkpoint, if so we update it.
-            if datetime.datetime.strptime(state, "%Y-%m-%dT%H:%M:%S.%f%z") < datetime.datetime.strptime(data['ModTime'], "%Y-%m-%dT%H:%M:%S.%f%z"):
-                state = data['ModTime']
+            try:
+                if datetime.datetime.strptime(state, "%Y-%m-%dT%H:%M:%S.%f%z") < datetime.datetime.strptime(data['ModTime'], "%Y-%m-%dT%H:%M:%S.%f%z"):
+                    state = data['ModTime']
+                    helper.log_debug(
+                        stanza_name + ' | ' + "Checkpoint key for alarm records was updated to " + state)
+            except:
                 helper.log_debug(
-                    stanza_name + ' | ' + "Checkpoint key for alarm records was updated to " + state)
+                        stanza_name + ' | ' + "Checkpoint key for alarm records was unable to be updated")
+                pass
         # Persist our checkpoint at the end of the audit records
         helper.save_check_point(account_name+'_last_alarm_record', state)
 
@@ -471,7 +481,7 @@ def collect_events(helper, ew):
         count = RESPONSE.json()['Count']
         helper.log_debug(stanza_name + ' | ' + "Found " +
                          str(count) + " inventory records to retrieve...")
-        results_per_page = 10  # adjust the number of results we pull per API call
+        results_per_page = 1  # adjust the number of results we pull per API call
         for i in range(0, count, results_per_page):
             RESPONSE = requests.request(
                 method='GET',
@@ -482,22 +492,23 @@ def collect_events(helper, ew):
                 verify=opt_validate_ssl,
                 proxies=r_proxy
             )
-
-            for data in RESPONSE.json()['Results']:
-                try:
+            try:
+                for data in RESPONSE.json()['Results']:
+                    try:
                     # remove things that just aren't helpful in splunk
-                    for thepop in ['Ancestors', 'Details', 'Owners', 'PermissionResources', 'RegisteredDevice']:
-                        data.pop(thepop)
-                    for thepop in ['ClassId', 'link']:
-                        data['ManagedObject'].pop(thepop)
-                except:
-                    pass
-                event = helper.new_event(
-                    source=account_name, index=index, sourcetype='cisco:intersight:condHclStatuses', data=json.dumps(data))
-                ew.write_event(event)
-                helper.log_debug(stanza_name + ' | ' +
+                        for thepop in ['Ancestors', 'Details', 'Owners', 'PermissionResources', 'RegisteredDevice']:
+                            data.pop(thepop)
+                        for thepop in ['ClassId', 'link']:
+                            data['ManagedObject'].pop(thepop)
+                    except:
+                        pass
+                    event = helper.new_event(
+                        source=account_name, index=index, sourcetype='cisco:intersight:condHclStatuses', data=json.dumps(data))
+                    ew.write_event(event)
+                    helper.log_debug(stanza_name + ' | ' +
                                  "Creating inventory for Moid "+data['Moid'])
-
+            except:
+                helper.log_debug(stanza_name + ' | ' + "HCL failed with code " + RESPONSE.json()['code'])
     else:
         helper.log_debug(stanza_name + ' | ' + "Skipping compute HCL status.")
 

splunkbase/releasenotes.md

@@ -24,3 +24,9 @@ Proxy support is configured at the Add-on level and not per-Input.  Click the `C
 Fixed an issue that could cause the add-on to fail if the Intersight SaaS account did not have any servers licensed at Essentials or higher when either the Compute inventory (because of the HCL feature) or Advisories Inventory items were selected.
 
 Fixed an issue that would cause the add-on to fail if Hyperflex inventory was selected and no Hyperflex clusters were present.
+
+## 1.2.7
+
+Worked around an issue that could cause the addon to fail if an Alarm or Audit Log event happen exactly at the second boundary and didn't leave any fraction of a second.  A more mathematically correct fix will have to wait for a future update.
+
+Worked around an issue that could cause the add-on to fail if the compute HCL query encountered a record that was associated with a server that's no longer licensed for the HCL feature.  This, too, will get a more thorough fix at some point in the future.