Changing constraints on cloud resource specifications tables to prevent extra rows

classes/OpenXdmod/Migration/Version900To950/DatabaseMigration.php

@@ -7,6 +7,7 @@
 
 use CCR\DB;
 use ETL\Utilities;
+use OpenXdmod\Setup\Console;
 
 /**
 * Migrate databases from version 9.0.0 to 9.5.0.
@@ -17,9 +18,34 @@ public function execute()
     {
         parent::execute();
 
+        $console = Console::factory();
+
         $dbh = DB::factory('datawarehouse');
         $mysql_helper = \CCR\DB\MySQLHelper::factory($dbh);
         if ($mysql_helper->tableExists('modw_cloud.cloud_resource_specs')) {
+
+            $staging_resource_sql = "SELECT
+                        COUNT(*)
+                    FROM
+                        modw_cloud.staging_resource_specifications
+                    GROUP BY
+                        resource_id, hostname, fact_date
+                    HAVING
+                        COUNT(*) > 1";
+
+            $staging_result = $dbh->query($staging_resource_sql);
+
+            if(count($staging_result) > 0) {
+                $console->displayMessage(<<<"EOT"
+This version of Open XDMoD changes the schema on two tables related to cloud utilization metrics. It appears that
+data in the table modw_cloud.staging_resource_specifications will violate these schema changes. The violation is that
+there cannot be two rows with the same resource ID, hostname, and date. Before you next ingest your cloud resource
+specification files you should either remove any extra rows or truncate this table and then shred and ingest all of
+your cloud resource specification files.
+EOT
+                );
+            }
+
             Utilities::runEtlPipeline(
                 ['cloud-migration-9_0_0-9_5_0'],
                 $this->logger,

configuration/etl/etl_tables.d/cloud_common/cloud_resource_specs.json

@@ -1,10 +1,10 @@
 {
-    "#": "Record types are global to all cloud resources",
+    "#": "Contains the start and end time for a specific set of vcpus and memory for a cloud host.",
 
     "table_definition": {
         "name": "cloud_resource_specs",
         "engine": "MyISAM",
-        "comment": "Record type: accounting, administrative, derived, etc.",
+        "comment": "Contains the start and end time for a specific set of vcpus and memory for a cloud host",
         "columns": [
             {
                 "name": "host_id",
@@ -62,20 +62,13 @@
                 "extra": "on update CURRENT_TIMESTAMP"
             }
         ],
-
-        "#": "Note that the primary key is (resource, type, identifier, create_time) rather than",
-        "#": "asset_id so we can take advantage of ON DUPLICATE KEY UPDATE to update existing entries",
-        "#": "and still use auto_increment for new entries.",
-
         "indexes": [
             {
                 "name": "PRIMARY",
                 "columns": [
                     "resource_id",
                     "hostname",
-                    "memory_mb",
-                    "vcpus",
-                    "start_date_ts"
+                    "start_day_id"
                 ],
                 "is_unique": true
             },

configuration/etl/etl_tables.d/cloud_common/staging_resource_specifications.json

@@ -1,10 +1,11 @@
 {
-    "#": "Record types are global to all cloud resources",
+    "#": "This table contains a specific set of vcpus and memory size for a host for",
+    "#": "a day. There should be no more than one record for a host on a day.",
 
     "table_definition": {
         "name": "staging_resource_specifications",
         "engine": "MyISAM",
-        "comment": "Record type: accounting, administrative, derived, etc.",
+        "comment": "Contains a specific set of vcpus and memory size for a host for a day",
         "columns": [
             {
                 "name": "hostname",
@@ -42,8 +43,6 @@
                 "columns": [
                     "resource_id",
                     "hostname",
-                    "memory_mb",
-                    "vcpus",
                     "fact_date"
                 ],
                 "is_unique": true