Monitoring agent for PostgreSQL 1.0 Fix Pack 12. Reference IBM

Transcription

1 Monitoring agent for PostgreSQL 1.0 Fix Pack 12 Reference IBM

2

3 Monitoring agent for PostgreSQL 1.0 Fix Pack 12 Reference IBM

4 Note Before using this information and the product it supports, read the information in Notices on page 35. This edition applies to version of the Monitoring agent for PostgreSQL and to all subsequent releases and modifications until otherwise indicated in new editions. Copyright IBM Corporation 2014, US Government Users Restricted Rights Use, duplication or disclosure restricted by GSA ADP Schedule Contract with IBM Corp.

5 Contents Chapter 1. Monitoring agent for PostgreSQL Chapter 2. Dashboards Default dashboards Group widgets for the Default dashboards Custom views Chapter 3. Eventing thresholds Predefined eventing thresholds Customized eventing thresholds Chapter 4. Attributes Data sets for the monitoring agent Attribute descriptions Availability data set Connection Num data set Database Query Counts data set Db Buff Hit Rate data set Db CPU data set Db Memory data set Db Num data set DB Size data set Lock Detail data set Locks data set Locks Filter data set Locks v96 data set Performance Object Status data set Server Info data set SQL Response Time data set SQL Statement Perf data set Take Action Status data set Thread Pool Status data set Accessibility features Notices Trademarks Terms and conditions for product documentation.. 37 IBM Online Privacy Statement Index Copyright IBM Corp. 2014, 2017 iii

6 iv Monitoring agent for PostgreSQL Reference

7 Chapter 1. Monitoring agent for PostgreSQL The Monitoring agent for PostgreSQL offers a central point of management for your PostgreSQL environment or application. The software provides a comprehensive means for gathering the information that is required to detect problems early and to prevent them. Information is standardized across the system. You can monitor multiple servers from a single console. By using the you can easily collect and analyze PostgreSQL specific information. Installing and configuring the agent Install the monitoring agent on the system where the application that you want to monitor is located. For more information, see the following documentation in IBM Knowledge Center: v IBM Performance Management v IBM Performance Management on Cloud Copyright IBM Corp. 2014,

8 2 Monitoring agent for PostgreSQL Reference

9 Chapter 2. Dashboards Open the Application Performance Dashboard in the Cloud APM console to see a status summary of all your applications. As you drill down to dashboards for specific applications and their supporting elements, more details are available about the selected item. Use the dashboards to proactively monitor your PostgreSQL deployment. Each dashboard page contains views with key performance indicators. When an application that includes PostgreSQL managed resources (also referred to as managed systems) is selected, the navigator and the Status Overview tab show PostgreSQL in the Components group: v Click Components to see a single PostgreSQL group widget that is displayed along with a group widget for every other data source type in the application. v Click the PostgreSQL subgroup to see a group widget for each managed resource in the application. v Click inside a PostgreSQL group widget or click a PostgreSQL managed resource from the navigator Instances section to open a dashboard with KPIs from the selected managed resource. For more information about the KPIs, click in the view or click in the dashboard banner. Default dashboards PostgreSQL Select the PostgreSQL managed component to drill down to more detailed metrics. Postgres Detail Dashboard Select a PostgreSQL managed instance to drill down to more detailed metrics. Additional group widgets These pop-up group widgets are displayed after you click a group widget for more details. Some group widgets have links to more granular information in a popup widget, described here. Group widgets for the Default dashboards PostgreSQL Select the PostgreSQL managed component to drill down to more detailed metrics. The following group widgets are available in this dashboard: PostgreSQL Status This group widget displays the PostgreSQL component summary and provides metrics to indicate the overall health of the PostgreSQL databases. The following KPIs for the group widget are derived from the attributes as described: v Active connections: The total number of active connections, which is the total number of rows in the pg_stat_activity table. This data is derived from the Connection Num attribute in the KPN CONNECTION NUM data set. v Current CPU used(%): The total CPU usage percentage, out of the system, of datbase processes ("postgres") since they started. This data is derived from the CPU Value attribute in the KPN DB CPU data set. v Disconnected DB: The total number of databases that cannot be connected to or are inactive. This data is derived from the Inactive Db Num attribute in the KPN DB NUM data set. Copyright IBM Corp. 2014,

10 v Normal DB: The total number of databases that can be connected to or are active. This data is derived from the Active Db Num attribute in the KPN DB NUM data set. v Physical memory used(mb): The memory usage of all database processes ("postgres") in MB. This data is derived from the Memory attribute in the KPN DB MEMORY data set. v Waiting locks: The number of locks on which transactions are waiting. This data is derived from the Lock Sum attribute in the KPN LOCKS FILTER data set. v Worst buffer hit rate(%): The database buffer hit percentage. This data is derived from the Buffer Hit Percentage attribute in the KPN DB BUFF HIT RATE data set. v Worst SQL execution time(ms): The execution time of the SQL statement in milliseconds. This data is derived from the Exec Time attribute in the KPN SQL STATEMENT PERF data set. Postgres Detail Dashboard Select a PostgreSQL managed instance to drill down to more detailed metrics. The following group widgets are available in this dashboard: Average Response Time by Operation(ms) - Top 10 This group widget displays the top 10 average response times of the SQL statements by database and operation. This widget includes the select, insert, update, and delete operations by database. The following KPIs for the group widget are derived from the attributes as described: v DB Name: The database name. This data is derived from the Db Name attribute in the KPN SQL RESPONSE TIME data set. v Response Time(ms): Is not derived directly from a single attribute. The value is calculated. Buffer Hit Rate - Lowest 5 Description needed in JSON The following KPIs for the group widget are derived from the attributes as described: v Active Connections: The number of active connections to the respective database. This data is derived from the Number Active Connections attribute in the KPN DB BUFF HIT RATE data set. v Allow Connections: Indicates whether the database allows connections. This data is derived from the Db Allow Conn attribute in the KPN DB BUFF HIT RATE data set. v DB Name: The name of the database. This data is derived from the Database Name attribute in the KPN DB BUFF HIT RATE data set. v Hit Rate(%): The database buffer hit percentage. Database Size - Top 5 This group widget displays the top five databases sorted by size in MB and includes all related tablespace names. The following KPIs for the group widget are derived from the attributes as described: v DB Name: The name of the database. This data is derived from the Database Name attribute in the KPN DB SIZE data set. v DB Size(MB): The size of the table space in MB. This data is derived from the Table Space Size MB attribute in the KPN DB SIZE data set. v Tablespace Used: The name of the table space. This data is derived from the Table Space Name attribute in the KPN DB SIZE data set. Request Number by Operation - Top 5 databases This group widget displays the top five execution counts of SQL statements by database. This widget includes the select, insert, update, and delete operations by database. 4 Monitoring agent for PostgreSQL Reference

11 The following KPIs for the group widget are derived from the attributes as described: v DB Name: The name of the database. This data is derived from the Database Name attribute in the KPN DATABASE QUERY COUNTS data set. v Deletes: The number of DELETE SQL statements that ran since the last pg_stat_statements module was enabled or reset. This data is derived from the Delete SQL Count attribute in the KPN DATABASE QUERY COUNTS data set. v Inserts: The number of INSERT SQL statements that ran since the last pg_stat_statements module was enabled or reset. This data is derived from the Insert SQL Count attribute in the KPN DATABASE QUERY COUNTS data set. v Selects: The number of SELECT SQL statements that ran since the last pg_stat_statements module was enabled or reset. This data is derived from the Select SQL Count attribute in the KPN DATABASE QUERY COUNTS data set. v Updates: The number of UPDATE SQL statements that ran since the last pg_stat_statements module was enabled or reset. This data is derived from the Update SQL Count attribute in the KPN DATABASE QUERY COUNTS data set. SQL Statement Execution Time - Slowest 5 This group widget displays the five slowest SQL statements according to execution time in milliseconds from all databases. The following KPIs for the group widget are derived from the attributes as described: v Affect Rows: Is not derived directly from a single attribute. The value is calculated. v Avg Time(ms): Is not derived directly from a single attribute. The value is calculated. v DB Name: The database name. This data is derived from the Database Name attribute in the KPN SQL STATEMENT PERF data set. v SQL Statement: Is not derived directly from a single attribute. The value is calculated. Waiting Lock Time - Top 5 This group widget displays the top five waiting lock times in seconds. The following KPIs for the group widget are derived from the attributes as described: v DB Name: The name of the database. This data is derived from the Database Name attribute in the KPN LOCK DETAIL data set. v Lock Mode: The lock mode to control concurrent access to data in tables. These modes can be used for application-controlled locking in situations where MVCC does not give the desired behavior. This data is derived from the Lock Mode attribute in the KPN LOCK DETAIL data set. v Waiting Pid: The process id of the transaction process that is waiting for a lock. This data is derived from the Wait PID attribute in the KPN LOCK DETAIL data set. v Waiting Time(s): Is not derived directly from a single attribute. The value is calculated. Additional group widgets These pop-up group widgets are displayed after you click a group widget for more details. Some group widgets have links to more granular information in a popup widget, described here. The following group widgets are available in this dashboard: Lock Detail This group widget displays information about the top five waiting lock times. Click any row to see a tooltip window that contains lock detail information. The following KPIs for the group widget are derived from the attributes as described: v DB name: The name of the database. This data is derived from the Database Name attribute in the KPN LOCK DETAIL data set. v Holding lock pid: The PID of the transcation process that is holding the lock. This data is derived from the Hold PID attribute in the KPN LOCK DETAIL data set. Chapter 2. Dashboards 5

12 Custom views v Holding lock sql stmt: The SQL Statement for the transaction that is holding the lock, limited to the first 256 characters of the query. This data is derived from the Hold SQL attribute in the KPN LOCK DETAIL data set. v ID: The object ID that is waiting for a lock, for example, if the lock type is tuple, it is the ID of tuple. This data is derived from the Lock Type ID attribute in the KPN LOCK DETAIL data set. v Lock mode: The lock mode to control concurrent access to data in tables. These modes can be used for application-controlled locking in situations where MVCC does not give the desired behavior. This data is derived from the Lock Mode attribute in the KPN LOCK DETAIL data set. v Lock type: The type of the lock object. This data is derived from the Lock Type attribute in the KPN LOCK DETAIL data set. v Waiting pid: The process id of the transaction process that is waiting for a lock. This data is derived from the Wait PID attribute in the KPN LOCK DETAIL data set. v Waiting time(s): The waiting time of the lock in seconds. This data is derived from the Wait Time attribute in the KPN LOCK DETAIL data set. After you select an application that includes a PostgreSQL managed resource, select the Custom Views tab to create and display custom dashboard pages with attribute values from the. You can quickly build monitoring pages for an application and save them for viewing by you alone or share with other users. 6 Monitoring agent for PostgreSQL Reference

13 Chapter 3. Eventing thresholds Thresholds test for certain conditions on your managed systems, such as memory usage over 95%, and raise an event when the conditions have been met. The agent comes with predefined thresholds that you can use to monitor your PostgreSQL environment. You can create additional thresholds for the areas of interest. After you click System Configuration > Threshold Manager, select PostgreSQL as the data source type to see all the available thresholds. Predefined eventing thresholds The thresholds are organized by the data set for which they were written. The has the following predefined thresholds: KPN_Active_Connections_Crit The number of active connections are greater than the critical threshold of 80 or there are no connections. The default configuration has the following SQL syntax: *IF *VALUE KPN_CONNECTION_NUM.Connection_Num *GT 80 *OR *VALUE KPN_CONNECTION_NUM.Connection_Num *EQ 0 This threshold is evaluated every 5 minutes. The severity of this threshold is Critical. KPN_Active_Connections_Warn The number of active connections are greater than the warning threshold of 60. The default configuration has the following SQL syntax: *IF *VALUE KPN_CONNECTION_NUM.Connection_Num *GT 60 *AND *VALUE KPN_CONNECTION_NUM.Connection_Num *LE 80 This threshold is evaluated every 5 minutes. The severity of this threshold is Warning. KPN_Buff_Hit_Percent_Crit The buffer hit rate of PostgreSQL is lower than 75 percent. The default configuration has the following SQL syntax: *IF *VALUE KPN_DB_BUFF_HIT_RATE.Buffer_Hit_Percentage *LT 75 *AND *VALUE KPN_DB_BUFF_HIT_RATE.Number_Active_Connections *NE 0 This threshold is evaluated every 5 minutes. The severity of this threshold is Critical. KPN_Buff_Hit_Percent_Warn The buffer hit rate of PostgreSQL is lower than 90 percent. The default configuration has the following SQL syntax: *IF *VALUE KPN_DB_BUFF_HIT_RATE.Buffer_Hit_Percentage *LT 90 *AND *VALUE KPN_DB_BUFF_HIT_RATE.Buffer_Hit_Percentage *GE 75 Copyright IBM Corp. 2014,

14 This threshold is evaluated every 5 minutes. The severity of this threshold is Warning. KPN_High_CPU_Crit The processor usage is greater than the critical threshold of 80 percent.the processor usage is calculated by adding the processor usage of all the processes that run on the processor cores. The processor usage can exceed 100% for multi-core processors. The default configuration has the following SQL syntax: *IF *VALUE KPN_DB_CPU.CPU_Value *GT 80 This threshold is evaluated every 5 minutes. The severity of this threshold is Critical. KPN_High_CPU_Warn The processor usage is greater than the warning threshold of 50 percent.the processor usage is calculated by adding the processor usage of all the processes that run on the processor cores. The processor usage can exceed 100% for multi-core processors. The default configuration has the following SQL syntax: *IF *VALUE KPN_DB_CPU.CPU_Value *GT 50 *AND *VALUE KPN_DB_CPU.CPU_Value *LE 80 This threshold is evaluated every 5 minutes. The severity of this threshold is Warning. KPN_Inactive_DBs_Warn Some databases cannot be accessed because of connection problem or the databases are inactive. The default configuration has the following SQL syntax: *IF *VALUE KPN_DB_NUM.Inactive_Db_Num *GT 0 This threshold is evaluated every 5 minutes. The severity of this threshold is Warning. KPN_Physical_Memory_High_Crit The physical memory that the database processes use is greater than the critical threshold of one GB. The default configuration has the following SQL syntax: *IF *VALUE KPN_DB_MEMORY.Memory *GT 1024 This threshold is evaluated every 5 minutes. The severity of this threshold is Critical. KPN_Physical_Memory_High_Warn The physical memory that the database processes use is greater than the warning threshold of 600 MB. The default configuration has the following SQL syntax: *IF *VALUE KPN_DB_MEMORY.Memory *GT 600 *AND *VALUE KPN_DB_MEMORY.Memory *LE 1024 This threshold is evaluated every 5 minutes. The severity of this threshold is Warning. KPN_PostgreSQL_Locks_Crit The number of waiting locks in PostgreSQL are greater than the critical threshold. The default configuration has the following SQL syntax: 8 Monitoring agent for PostgreSQL Reference

15 *IF *VALUE KPN_LOCKS_FILTER.Lock_Sum *GT 5 This threshold is evaluated every 5 minutes. The severity of this threshold is Critical. KPN_PostgreSQL_Locks_Warn The number of waiting locks in PostgreSQL are greater than the warning threshold. The default configuration has the following SQL syntax: *IF *VALUE KPN_LOCKS_FILTER.Lock_Sum *GT 0 *AND *VALUE KPN_LOCKS_FILTER.Lock_Sum *LE 5 This threshold is evaluated every 5 minutes. The severity of this threshold is Warning. KPN_SQL_Statement_Perf_Crit The execution time of the SQL statement is greater than five seconds. The default configuration has the following SQL syntax: *IF *VALUE KPN_SQL_STATEMENT_PERF.Exec_Time *GT 5000 This threshold is evaluated every 5 minutes. The severity of this threshold is Critical. KPN_SQL_Statement_Perf_Warn The execution time of the SQL statement is greater than one second. The default configuration has the following SQL syntax: *IF *VALUE KPN_SQL_STATEMENT_PERF.Exec_Time *GT 1000 *AND *VALUE KPN_SQL_STATEMENT_PERF.Exec_Time *LE 5000 This threshold is evaluated every 5 minutes. The severity of this threshold is Warning. Customized eventing thresholds You can use the predefined eventing thresholds as a starting point for event monitoring, and create your own eventing thresholds as conditions arise that you want to monitor. Hover your mouse over a help icon to get user assistance for the field. Click the help tool to open the help system for the Threshold Manager help topic. The has many data sets that you can use to create eventing thresholds to monitor for specific conditions. Tip: If you are an IBM Tivoli Monitoring user, data sets are equivalent to attribute groups. Chapter 3. Eventing thresholds 9

16 10 Monitoring agent for PostgreSQL Reference

17 Chapter 4. Attributes Attributes are the application properties that are being measured and reported by the Monitoring agent for PostgreSQL. Attributes make up the key performance indicators (KPIs) that are reported, and you can use them to create thresholds for conditions that you want to monitor. About attributes Attributes are organized into data sets (also referred to as attribute groups). The values can be selectively displayed in dashboards or used to define a threshold. When you view dashboards and use thresholds, data samples of the attributes in the data set are taken. Dashboards Attributes are displayed in the dashboards. The widgets specify which attribute values to request from a monitoring agent. You can edit the Components group widgets that display in the Application Performance Dashboard to control how the group widgets are displayed, their positioning, and to adjust the summary thresholds. After you select a defined application, select the group that is named Components, and click Actions > Edit to open the editor. The Attribute Details tab is displayed after you select an instance from the navigator Instances section (renamed to the selected subgroup name). Alternatively, you can click inside a summary group widget in the group or subgroup dashboard page. Use the tools in the Attribute Details tab to create custom charts and tables with data samples from the available data sets. Thresholds You can define thresholds that monitor the state of your operating system, database, or application and open an event when the threshold is exceeded. You use attributes to define thresholds that describe a condition that you want to test. After the threshold is started, the values that are specified in the threshold definition are compared with the values that are collected by the. After the condition is met, an event is registered and you are alerted by indicator icons in the Application Performance Dashboard navigator and the Events tab. The comes with predefined thresholds that are enabled and started with the agent. If you edit a predefined threshold definition, such as to change a condition or severity, it is no longer treated as a predefined threshold but considered a custom threshold. When the sampled values for the attributes that are referenced in a threshold meet the condition (such as attribute A is greater than 80%), an event is opened. Note the following conditions: v v v When no data can be collected for a data set, an empty result is returned (no rows of data) When a specific attribute cannot be collected, the value 0 or is returned unless otherwise specified in a particular attribute (for example, N/A ) Any numeric attribute value that is greater than the largest (positive or negative) number that can be represented by that type returns the corresponding maximum or minimum value. For example, the maximum value for a 32-bit number is 2,147,483,647. These values are displayed as text values that are defined by the data set, such as Value Exceeds Maximum or Value Exceeds Minimum. Additional information about attributes For a list of the data sets, a list of the attributes in each data set, and descriptions of the attributes in the, see Data sets for the monitoring agent on page 12 and Attribute descriptions on page 13. Copyright IBM Corp. 2014,

18 Data sets for the monitoring agent The contains the following data sets. v Data set name: Availability Table name: KPNAVAIL Historical table name: KPN_AVAILABILITY or KPNAVAIL v Data set name: Connection Num Table name: KPNCONNSUM Historical table name: KPN_CONNECTION_NUM or KPNCONNSUM v Data set name: Database Query Counts Table name: KPNDATABAS Historical table name: KPN_DATABASE_QUERY_COUNTS or KPNDATABAS v Data set name: Db Buff Hit Rate Table name: KPNDBBUFFH Historical table name: KPN_DB_BUFF_HIT_RATE or KPNDBBUFFH v Data set name: Db CPU Table name: KPNDBCPU Historical table name: KPN_DB_CPU or KPNDBCPU v Data set name: Db Memory Table name: KPNDBMEM Historical table name: KPN_DB_MEMORY or KPNDBMEM v Data set name: Db Num Table name: KPNDBNUM Historical table name: KPN_DB_NUM or KPNDBNUM v Data set name: DB Size Table name: KPNDBSIZE Historical table name: KPN_DB_SIZE or KPNDBSIZE v Data set name: Lock Detail Table name: KPNLOCKDET Historical table name: KPN_LOCK_DETAIL or KPNLOCKDET v Data set name: Locks Table name: KPNLOCKS Historical table name: KPN_LOCKS v Data set name: Locks Filter Table name: KPNLOCKSF Historical table name: KPN_LOCKS_FILTER or KPNLOCKSF v Data set name: Locks v96 Table name: KPNLOCKSN Historical table name: KPN_LOCKS_V96 or KPNLOCKSN v Data set name: Performance Object Status Table name: KPNPOBJST Historical table name: KPN_PERFORMANCE_OBJECT_STATUS or KPNPOBJST v Data set name: Server Info Table name: KPNSVRINFO Historical table name: KPN_SERVER_INFO or KPNSVRINFO 12 Monitoring agent for PostgreSQL Reference

19 v Data set name: SQL Response Time Table name: KPNSQLRESP Historical table name: KPN_SQL_RESPONSE_TIME or KPNSQLRESP v Data set name: SQL Statement Perf Table name: KPNSQLPERF Historical table name: KPN_SQL_STATEMENT_PERF or KPNSQLPERF v Data set name: Take Action Status Table name: KPNTACTST Historical table name: KPN_TAKE_ACTION_STATUS or KPNTACTST v Data set name: Thread Pool Status Table name: KPNTHPLST Historical table name: KPN_THREAD_POOL_STATUS or KPNTHPLST Attribute descriptions Attributes in each data set collect data that the agent uses for monitoring. The descriptions of the data sets contain information such as description, type, and names for each attribute in the data set. Some attributes are designated as key attributes. A key attribute is an identifier attribute for the data set. Availability data set The Availability data set contains the availability data for all processes and services that make up this application. This data set contains the following attributes: Application Component attribute The descriptive name of a part of the application. This attribute is a key attribute. The type is string. The following names are defined for this attribute: APPLICATION_COMPONENT or COMPONENT (historical name), Application Component (caption), Application_Component (attribute name), and COMPONENT (column name). Command Line attribute The program name and any arguments specified on the command line when the process was started. For Service or Functionality test, this attribute has the value N/A. The type is string with enumerated values. The following values are defined: N/A (N/A). Any value that does not have a definition here is displayed in the User Interface. The following names are defined for this attribute: COMMAND_LINE or CMDLINE (historical name), Command Line (caption), Command_Line (attribute name), and CMDLINE (column name). Full Name attribute The full name of the process that includes the path. The type is string with enumerated values. The following values are defined: N/A (N/A). Any value that does not have a definition here is displayed in the User Interface. The following names are defined for this attribute: FULL_NAME or FULLNAME (historical name), Full Name (caption), Full_Name (attribute name), and FULLNAME (column name). Functionality Test Message attribute The text message that corresponds to the Functionality Test Status. This attribute is only valid for Chapter 4. Attributes 13

20 functionality tests. The type is string with enumerated values. The following values are defined: N/A (N/A). Any value that does not have a definition here is displayed in the User Interface. The following names are defined for this attribute: FUNCTIONALITY_TEST_MESSAGE or FUNCMSG (historical name), Functionality Test Message (caption), Functionality_Test_Message (attribute name), and FUNCMSG (column name). Functionality Test Status attribute The return code of the functionality test. When the monitored application is running correctly, 'SUCCESS' is displayed. 'NOT_RUNNING' is displayed when it is not running correctly. 'N/A' is displayed when the row does not represent a functionality test. The type is integer with enumerated values. The following values are defined: SUCCESS (0), N/A (1), GENERAL ERROR (2), WARNING (3), NOT RUNNING (4), DEPENDENT NOT RUNNING (5), ALREADY RUNNING (6), PREREQ NOT RUNNING (7), TIMED OUT (8), DOESNT EXIST (9), UNKNOWN (10), DEPENDENT STILL RUNNING (11), INSUFFICIENT USER AUTHORITY (12). Any value that does not have a definition here is displayed in the User Interface. The following names are defined for this attribute: FUNCTIONALITY_TEST_STATUS or FUNCSTATUS (historical name), Functionality Test Status (caption), Functionality_Test_Status (attribute name), and FUNCSTATUS (column name). Name attribute The name of the process, service, or functionality test. This name matches the executable name of the process, the service short name or the name of the process used to test the application. The type is string with enumerated values. The following values are defined: N/A (N/A). Any value that does not have a definition here is displayed in the User Interface. The following names are defined for this attribute: NAME (historical name), Name (caption), Name (attribute name), and NAME (column name). Node attribute The managed system name of the agent. This attribute is a key attribute. The type is string. The following names are defined for this attribute: NODE (historical name), Node (caption), ORIGINNODE (attribute name), and ORIGINNODE (column name). Page Faults per Sec attribute The rate of page faults for the process measured in faults per second. This attribute contains only valid data for processes. The type is integer (32-bit gauge). The following names are defined for this attribute: PAGE_FAULTS_PER_SEC or PAGEFAULTS (historical name), Page Faults per Sec (caption), Page_Faults_per_Sec (attribute name), and PAGEFAULTS (column name). Percent Privileged Time attribute The percentage of the available CPU time being used by this process for privileged operation. The type is integer (32-bit gauge). The following names are defined for this attribute: PERCENT_PRIVILEGED_TIME or PERCPRIV (historical name), Percent Privileged Time (caption), Percent_Privileged_Time (attribute name), and PERCPRIV (column name). Percent Processor Time attribute The percentage of the elapsed time that this process used the processor to execute instructions. The type is integer (32-bit gauge). The following names are defined for this attribute: PERCENT_PROCESSOR_TIME or PERCPROC (historical name), Percent Processor Time (caption), Percent_Processor_Time (attribute name), and PERCPROC (column name). Percent User Mode Time attribute The percentage of the available CPU time being used by this process for user mode operation. The type is integer (32-bit gauge). 14 Monitoring agent for PostgreSQL Reference

21 The following names are defined for this attribute: PERCENT_USER_MODE_TIME or PERCUSER (historical name), Percent User Mode Time (caption), Percent_User_Mode_Time (attribute name), and PERCUSER (column name). PID attribute The process ID associated with the process. This attribute contains only valid data for processes. The type is integer (32-bit gauge). The following names are defined for this attribute: PID (historical name), PID (caption), PID (attribute name), and PID (column name). Status attribute The status of the application component. v v v For processes 'UP', 'DOWN', 'WARNING', or 'PROCESS_DATA_NOT_AVAILABLE': 'PROCESS_DATA_NOT_AVAILABLE' is displayed for a process when the matching process is running but the resource use information cannot be collected for that process. For services 'UP', 'DOWN', or 'UNKNOWN': 'UNKNOWN' is displayed when the service is not installed. For functionality tests: 'PASSED' or 'FAILED' is displayed. The type is integer with enumerated values. The following values are defined: DOWN (0), UP (1), WARNING (2), UNKNOWN (3), PASSED (4), FAILED (5), PROCESS DATA NOT AVAILABLE (6). Any value that does not have a definition here is displayed in the User Interface. The following names are defined for this attribute: STATUS (historical name), Status (caption), Status (attribute name), and STATUS (column name). Thread Count attribute The number of threads currently allocated by this process. This attribute contains only valid data for processes. The type is integer (32-bit gauge). The following names are defined for this attribute: THREAD_COUNT or THREADS (historical name), Thread Count (caption), Thread_Count (attribute name), and THREADS (column name). Timestamp attribute The local time at the agent when the data was collected. The type is string. The following names are defined for this attribute: TIMESTAMP (historical name), Timestamp (caption), Timestamp (attribute name), and TIMESTAMP (column name). Type attribute The type of the application component. Components are processes, services, or functionality tests. The type is integer with enumerated values. The following values are defined: PROCESS (0), SERVICE (1), FUNCTIONALITY TEST (2). Any value that does not have a definition here is displayed in the User Interface. The following names are defined for this attribute: TYPE (historical name), Type (caption), Type (attribute name), and TYPE (column name). Virtual Size attribute The virtual size (in MB) of the process. The type is integer (32-bit gauge). The following names are defined for this attribute: VIRTUAL_SIZE or VIRTSIZE (historical name), Virtual Size (caption), Virtual_Size (attribute name), and VIRTSIZE (column name). Working Set Size attribute The working set size of the process in MB. This attribute contains only valid data for processes. The type is integer (32-bit gauge). The following names are defined for this attribute: WORKING_SET_SIZE or WORKSET (historical name), Working Set Size (caption), Working_Set_Size (attribute name), and WORKSET (column name). Chapter 4. Attributes 15

22 Connection Num data set The data set provides the number of active connections. This data set contains the following attributes: Connection Num attribute The total number of active connections, which is the total number of rows in the pg_stat_activity table. This attribute is a key attribute. The type is integer (64-bit gauge). The following names are defined for this attribute: CONNECTION_NUM or CONNNUM (historical name), Connection Num (caption), Connection_Num (attribute name), and CONNNUM (column name). Node attribute The managed system name of the agent. This attribute is a key attribute. The type is string. The following names are defined for this attribute: NODE (historical name), Node (caption), ORIGINNODE (attribute name), and ORIGINNODE (column name). Timestamp attribute The local time at the agent when the data was collected. The type is string. The following names are defined for this attribute: TIMESTAMP (historical name), Timestamp (caption), Timestamp (attribute name), and TIMESTAMP (column name). Database Query Counts data set The execution count of SQL statements by the respective databases. This data set contains the following attributes: Database Name attribute The name of the database. This attribute is a key attribute. The type is string. The following names are defined for this attribute: DATABASE_NAME or DBNAME (historical name), Database Name (caption), Database_Name (attribute name), and DBNAME (column name). Delete SQL Count attribute The number of DELETE SQL statements that ran since the last pg_stat_statements module was enabled or reset. The type is integer with enumerated values. The following values are defined: No value (-1). Any value that does not have a definition here is displayed in the User Interface. The following names are defined for this attribute: DELETE_SQL_COUNT or DELETE_SQL (historical name), Delete SQL Count (caption), Delete_SQL_Count (attribute name), and DELETE_SQL (column name). Insert SQL Count attribute The number of INSERT SQL statements that ran since the last pg_stat_statements module was enabled or reset. The type is integer with enumerated values. The following values are defined: No value (-1). Any value that does not have a definition here is displayed in the User Interface. The following names are defined for this attribute: INSERT_SQL_COUNT or INSERT_SQL (historical name), Insert SQL Count (caption), Insert_SQL_Count (attribute name), and INSERT_SQL (column name). Node attribute The managed system name of the agent. This attribute is a key attribute. The type is string. The following names are defined for this attribute: NODE (historical name), Node (caption), ORIGINNODE (attribute name), and ORIGINNODE (column name). 16 Monitoring agent for PostgreSQL Reference

23 Select SQL Count attribute The number of SELECT SQL statements that ran since the last pg_stat_statements module was enabled or reset. The type is integer with enumerated values. The following values are defined: No value (-1). Any value that does not have a definition here is displayed in the User Interface. The following names are defined for this attribute: SELECT_SQL_COUNT or SELECT_SQL (historical name), Select SQL Count (caption), Select_SQL_Count (attribute name), and SELECT_SQL (column name). Timestamp attribute The local time at the agent when the data was collected. The type is string. The following names are defined for this attribute: TIMESTAMP (historical name), Timestamp (caption), Timestamp (attribute name), and TIMESTAMP (column name). Total Count attribute The number of SQL statements that ran since the last pg_stat_statements module was enabled or reset. The type is integer (64-bit gauge). The following names are defined for this attribute: TOTAL_COUNT or TOTAL_PER_ (historical name), Total Count (caption), Total_Count (attribute name), and TOTAL_PER_ (column name). Update SQL Count attribute The number of UPDATE SQL statements that ran since the last pg_stat_statements module was enabled or reset. The type is integer with enumerated values. The following values are defined: No value (-1). Any value that does not have a definition here is displayed in the User Interface. The following names are defined for this attribute: UPDATE_SQL_COUNT or UPDATE_SQL (historical name), Update SQL Count (caption), Update_SQL_Count (attribute name), and UPDATE_SQL (column name). Db Buff Hit Rate data set The buffer hit percentage data set provides information about the hit percentage of all buffers that are available for the database. This data set contains the following attributes: Buffer Hit Percent attribute The database buffer hit percentage. 101 represents 'No active connections' The type is real number (32-bit gauge) with two decimal places of precision. The following names are defined for this attribute: BUFFER_HIT_PERCENT or BUFFER_HIT (historical name), Buffer Hit Percent (caption), Buffer_Hit_Percent (attribute name), and BUFFER_HIT (column name). Buffer Hit Percentage attribute The database buffer hit percentage. The type is real number (32-bit gauge) with two decimal places of precision with enumerated values. The following values are defined: No active connections (101). Any value that does not have a definition here is displayed in the User Interface. The following names are defined for this attribute: BUFFER_HIT_PERCENTAGE or BUFFERHITP (historical name), Buffer Hit Percentage (caption), Buffer_Hit_Percentage (attribute name), and BUFFERHITP (column name). Database Name attribute The name of the database. This attribute is a key attribute. The type is string. The following names are defined for this attribute: DATABASE_NAME or DBNAME (historical name), Database Name (caption), Database_Name (attribute name), and DBNAME (column name). Chapter 4. Attributes 17

24 Db Allow Conn attribute Indicates whether the database allows connections. The type is string with enumerated values. The following values are defined: true (true), false (false). Any value that does not have a definition here is displayed in the User Interface. The following names are defined for this attribute: DB_ALLOW_CONN or DBAWCONN (historical name), Db Allow Conn (caption), Db_Allow_Conn (attribute name), and DBAWCONN (column name). Node attribute The managed system name of the agent. This attribute is a key attribute. The type is string. The following names are defined for this attribute: NODE (historical name), Node (caption), ORIGINNODE (attribute name), and ORIGINNODE (column name). Number Active Connections attribute The number of active connections to the respective database. The type is integer (32-bit gauge). The following names are defined for this attribute: NUMBER_ACTIVE_CONNECTIONS or NUMBER_ACT (historical name), Number Active Connections (caption), Number_Active_Connections (attribute name), and NUMBER_ACT (column name). Timestamp attribute The local time at the agent when the data was collected. The type is string. The following names are defined for this attribute: TIMESTAMP (historical name), Timestamp (caption), Timestamp (attribute name), and TIMESTAMP (column name). Db CPU data set The total CPU usage percentage, out of the system, of database processes ("postgres") since they started. This data set contains the following attributes: CPU Value attribute The total CPU usage percentage, out of the system, of datbase processes ("postgres") since they started. This attribute is a key attribute. The type is real number (32-bit gauge) with one decimal places of precision. The following names are defined for this attribute: CPU_VALUE or CPUVALUE (historical name), CPU Value (caption), CPU_Value (attribute name), and CPUVALUE (column name). Node attribute The managed system name of the agent. This attribute is a key attribute. The type is string. The following names are defined for this attribute: NODE (historical name), Node (caption), ORIGINNODE (attribute name), and ORIGINNODE (column name). Timestamp attribute The local time at the agent when the data was collected. The type is string. The following names are defined for this attribute: TIMESTAMP (historical name), Timestamp (caption), Timestamp (attribute name), and TIMESTAMP (column name). Db Memory data set The memory usage of all database processes ("postgres") in MB. This data set contains the following attributes: Memory attribute The memory usage of all database processes ("postgres") in MB. This attribute is a key attribute. The type is real number (32-bit gauge) with one decimal places of precision. 18 Monitoring agent for PostgreSQL Reference

25 The following names are defined for this attribute: MEMORY (historical name), Memory (caption), Memory (attribute name), and MEMORY (column name). Node attribute The managed system name of the agent. This attribute is a key attribute. The type is string. The following names are defined for this attribute: NODE (historical name), Node (caption), ORIGINNODE (attribute name), and ORIGINNODE (column name). Timestamp attribute The local time at the agent when the data was collected. The type is string. The following names are defined for this attribute: TIMESTAMP (historical name), Timestamp (caption), Timestamp (attribute name), and TIMESTAMP (column name). Db Num data set The total number of databases, both inactive and active. This data set contains the following attributes: Active Db Num attribute The total number of databases that can be connected to or are active. This attribute is a key attribute. The type is integer (32-bit gauge). The following names are defined for this attribute: ACTIVE_DB_NUM or DBNUMAC (historical name), Active Db Num (caption), Active_Db_Num (attribute name), and DBNUMAC (column name). Inactive Db Num attribute The total number of databases that cannot be connected to or are inactive. This attribute is a key attribute. The type is integer (32-bit gauge). The following names are defined for this attribute: INACTIVE_DB_NUM or DBNUMIN (historical name), Inactive Db Num (caption), Inactive_Db_Num (attribute name), and DBNUMIN (column name). Node attribute The managed system name of the agent. This attribute is a key attribute. The type is string. The following names are defined for this attribute: NODE (historical name), Node (caption), ORIGINNODE (attribute name), and ORIGINNODE (column name). Timestamp attribute The local time at the agent when the data was collected. The type is string. The following names are defined for this attribute: TIMESTAMP (historical name), Timestamp (caption), Timestamp (attribute name), and TIMESTAMP (column name). DB Size data set Data size information for respective databases in MB. This data set contains the following attributes: Database Name attribute The name of the database. This attribute is a key attribute. The type is string. The following names are defined for this attribute: DATABASE_NAME or DBNAME (historical name), Database Name (caption), Database_Name (attribute name), and DBNAME (column name). Node attribute The managed system name of the agent. This attribute is a key attribute. The type is string. Chapter 4. Attributes 19

26 The following names are defined for this attribute: NODE (historical name), Node (caption), ORIGINNODE (attribute name), and ORIGINNODE (column name). Table Space Name attribute The name of the table space. This attribute is a key attribute. The type is string. The following names are defined for this attribute: TABLE_SPACE_NAME or TSNAME (historical name), Table Space Name (caption), Table_Space_Name (attribute name), and TSNAME (column name). Table Space Size MB attribute The size of the table space in MB. The type is integer (32-bit gauge). The following names are defined for this attribute: TABLE_SPACE_SIZE_MB or TSSIZE (historical name), Table Space Size MB (caption), Table_Space_Size_MB (attribute name), and TSSIZE (column name). Timestamp attribute The local time at the agent when the data was collected. The type is string. The following names are defined for this attribute: TIMESTAMP (historical name), Timestamp (caption), Timestamp (attribute name), and TIMESTAMP (column name). Lock Detail data set The data set provides detailed information about the database locks. This data set contains the following attributes: Database Name attribute The name of the database. This attribute is a key attribute. The type is string. The following names are defined for this attribute: DATABASE_NAME or DBNAME (historical name), Database Name (caption), Database_Name (attribute name), and DBNAME (column name). Hold App Name attribute The name of the application that is holding the lock. The type is string. The following names are defined for this attribute: HOLD_APP_NAME or HOLDAPPNAM (historical name), Hold App Name (caption), Hold_App_Name (attribute name), and HOLDAPPNAM (column name). Hold PID attribute The PID of the transcation process that is holding the lock. This attribute is a key attribute. The type is integer (32-bit numeric property). The following names are defined for this attribute: HOLD_PID or HOLDPID (historical name), Hold PID (caption), Hold_PID (attribute name), and HOLDPID (column name). Hold SQL attribute The SQL Statement for the transaction that is holding the lock, limited to the first 256 characters of the query. The type is string. The following names are defined for this attribute: HOLD_SQL or HOLDSQL (historical name), Hold SQL (caption), Hold_SQL (attribute name), and HOLDSQL (column name). Lock Mode attribute The lock mode to control concurrent access to data in tables. These modes can be used for application-controlled locking in situations where MVCC does not give the desired behavior. The type is string. The following names are defined for this attribute: LOCK_MODE or LOCKMODE (historical name), Lock Mode (caption), Lock_Mode (attribute name), and LOCKMODE (column name). 20 Monitoring agent for PostgreSQL Reference

27 Lock Type attribute The type of the lock object. The type is string. The following names are defined for this attribute: LOCK_TYPE or LOCKTYPE (historical name), Lock Type (caption), Lock_Type (attribute name), and LOCKTYPE (column name). Lock Type ID attribute The object ID that is waiting for a lock, for example, if the lock type is tuple, it is the ID of tuple. The type is string. The following names are defined for this attribute: LOCK_TYPE_ID or LOCKTYPEID (historical name), Lock Type ID (caption), Lock_Type_ID (attribute name), and LOCKTYPEID (column name). Node attribute The managed system name of the agent. This attribute is a key attribute. The type is string. The following names are defined for this attribute: NODE (historical name), Node (caption), ORIGINNODE (attribute name), and ORIGINNODE (column name). Timestamp attribute The local time at the agent when the data was collected. The type is string. The following names are defined for this attribute: TIMESTAMP (historical name), Timestamp (caption), Timestamp (attribute name), and TIMESTAMP (column name). Wait App Name attribute The name of the application that is waiting for the lock. The type is string. The following names are defined for this attribute: WAIT_APP_NAME or WAITAPPNAM (historical name), Wait App Name (caption), Wait_App_Name (attribute name), and WAITAPPNAM (column name). Wait PID attribute The process id of the transaction process that is waiting for a lock. This attribute is a key attribute. The type is integer (32-bit numeric property). The following names are defined for this attribute: WAIT_PID or WAITPID (historical name), Wait PID (caption), Wait_PID (attribute name), and WAITPID (column name). Wait Time attribute The waiting time of the lock in seconds. The type is string. The following names are defined for this attribute: WAIT_TIME or WAITTIME (historical name), Wait Time (caption), Wait_Time (attribute name), and WAITTIME (column name). Locks data set The number of locks on which one or more transactions are waiting. This data set contains the following attributes: Lock Sum attribute The number of locks on which transactions are waiting. This attribute is a key attribute. The type is integer (64-bit gauge). The following names are defined for this attribute: LOCK_SUM or LOCKSUM (historical name), Lock Sum (caption), Lock_Sum (attribute name), and LOCKSUM (column name). Node attribute The managed system name of the agent. This attribute is a key attribute. The type is string. The following names are defined for this attribute: NODE (historical name), Node (caption), ORIGINNODE (attribute name), and ORIGINNODE (column name). Chapter 4. Attributes 21

28 Timestamp attribute The local time at the agent when the data was collected. The type is string. The following names are defined for this attribute: TIMESTAMP (historical name), Timestamp (caption), Timestamp (attribute name), and TIMESTAMP (column name). Locks Filter data set The number of locks on which one or more transactions are waiting. This data set contains the following attributes: Lock Sum attribute The number of locks on which transactions are waiting. This attribute is a key attribute. The type is integer (64-bit gauge). The following names are defined for this attribute: LOCK_SUM or LOCKSUM (historical name), Lock Sum (caption), Lock_Sum (attribute name), and LOCKSUM (column name). Node attribute The managed system name of the agent. This attribute is a key attribute. The type is string. The following names are defined for this attribute: NODE (historical name), Node (caption), ORIGINNODE (attribute name), and ORIGINNODE (column name). Timestamp attribute The local time at the agent when the data was collected. The type is string. The following names are defined for this attribute: TIMESTAMP (historical name), Timestamp (caption), Timestamp (attribute name), and TIMESTAMP (column name). Locks v96 data set The number of locks on which one or more transactions are waiting. This data set contains the following attributes: Lock Sum attribute The number of locks on which transactions are waiting. This attribute is a key attribute. The type is integer (64-bit gauge). The following names are defined for this attribute: LOCK_SUM or LOCKSUM (historical name), Lock Sum (caption), Lock_Sum (attribute name), and LOCKSUM (column name). Node attribute The managed system name of the agent. This attribute is a key attribute. The type is string. The following names are defined for this attribute: NODE (historical name), Node (caption), ORIGINNODE (attribute name), and ORIGINNODE (column name). Timestamp attribute The local time at the agent when the data was collected. The type is string. The following names are defined for this attribute: TIMESTAMP (historical name), Timestamp (caption), Timestamp (attribute name), and TIMESTAMP (column name). Performance Object Status data set The Performance Object Status data set contains information that reflects the status of other data sets so you can see the status of all performance objects that make up this application all at once. Each of these other performance data sets is represented by a row in this table (or other type of view). The status for a data set reflects the result of the last attempt to collect data for that data set, so you can see whether the 22 Monitoring agent for PostgreSQL Reference