Enable/Disable Oracle Cluster Health Advisor monitoring using chactl

chad stands for the Cluster Health Advisor (CHA) daemon which is is part of the Oracle Autonomous Health Framework(AHF), It continuously monitors cluster nodes and Oracle RAC databases for performance and availability issues. 

Check ochad status:

[grid@dbracnode1 ~]$  crsctl stat res ora.chad -t

--------------------------------------------------------------------------------

Name           Target  State        Server                   State details

--------------------------------------------------------------------------------

Local Resources

--------------------------------------------------------------------------------

ora.chad

               OFFLINE  OFFLINE       dbracnode1        STABLE

               OFFLINE  OFFLINE       dbracnode2        STABLE

--------------------------------------------------------------------------------               

[grid@dbracnode1 ~]$ crsctl status res ora.chad

NAME=ora.chad

TYPE=ora.chad.type

TARGET=OFFLINE, OFFLINE

STATE=OFFLINE, OFFLINE

[grid@dbracnode1 ~]$ chactl status

not monitoring nodes

not monitoring databases

we can start ochad using srvctl 

[grid@dbracnode1 ~]$ srvctl start cha

check status again, it's online now 

[grid@dbracnode1 ~]$ crsctl status res ora.chad

NAME=ora.chad

TYPE=ora.chad.type

TARGET=ONLINE             , ONLINE

STATE=ONLINE on dbracnode1, ONLINE on dbracnode2

[grid@dbracnode1 ~]$  crsctl stat res ora.chad -t

--------------------------------------------------------------------------------

Name           Target  State        Server                   State details

--------------------------------------------------------------------------------

Local Resources

--------------------------------------------------------------------------------

ora.chad

               ONLINE  ONLINE       dbracnode1        STABLE

               ONLINE  ONLINE       dbracnode2        STABLE

--------------------------------------------------------------------------------

[grid@dbracnode1 ~]$

    

we can use chactl to manage cluster health advisor 

check status:            

[grid@dbracnode1 ~]$ chactl status

monitoring nodes dbracnode1, dbracnode2

not monitoring databases

[grid@dbracnode1 ~]$ chactl status database -db tstdb01 -verbose

not monitoring database tstdb01

add db to cha monitoring:

[grid@dbracnode1 ~]$ chactl monitor database -db tstdb01

Now db added to cha monitoring 

[grid@dbracnode1 ~]$ chactl status cluster -verbose

monitoring nodes dbracnode1, dbracnode2 using model DEFAULT_CLUSTER

[grid@dbracnode1 ~]$ chactl status database -db tstdb01 -verbose

monitoring database tstdb01 using model DEFAULT_DB

monitoring instance tstdb011 on node dbracnode1

monitoring instance tstdb012 on node dbracnode2

remove db from cha monitoring:

[grid@dbracnode1 ~]$ chactl unmonitor database -db tstdb01

[grid@dbracnode1 ~]$ chactl status database -db tstdb01 -verbose

not monitoring database tstdb01

chactl useful commands:

[grid@dbracnode1 ~]$ chactl -help

Usage:

chactl monitor cluster [-model <model_name> [-force]]

chactl monitor database -db <db_unique_name>

                        [-model <model_name> [-force]]

chactl unmonitor database -db <db_unique_name>

chactl status [cluster|database [-db <db_unique_name>]] [-verbose]

chactl config [cluster|database -db <dbname>]

chactl calibrate {cluster|database -db <db_unique_name>}

                 -model <model_name> [-force]

                 [-timeranges "start=<time_stamp>,end=<time_stamp>,..."]

                 [-kpiset "name=<kpi_name> min=<val> max=<val>,..."]

    WHERE:

        -timeranges "start=<time_stamp>,end=<time_stamp>,..." :

             Timestamp must be in format 'YYYY-MM-DD HH24:MI:SS'

    KPI for db:

        CPUPERCENT - CPU utilization - Percent

        IOREAD - Disk read - Mbyte/sec

        DBTIMEPERCALL - Database time per user call - usec/call

        IOWRITE - Disk write - Mbyte/sec

        IOTHROUGHPUT - Disk throughput - IO/sec

    KPI for cluster:

        CPUPERCENT - CPU utilization - Percent

        IOREAD - Disk read - Mbyte/sec

        IOWRITE - Disk write - Mbyte/sec

        IOTHROUGHPUT - Disk throughput - IO/sec

chactl query diagnosis [-cluster|-db <db_uniq_name>]

                       [-start <time> -end <time>]

                       [-htmlfile <file_name>]

chactl query model [-name <model_name> [-verbose]]

chactl query repository

chactl query calibration {-cluster|-db <db_uniq_name>}

        [-timeranges "start=<time_stamp>,end=<time_stamp>,..."]

        [-kpiset "name=<kpi_name> min=<val> max=<val>,..."]

        [-interval <val>]

    WHERE:

        -interval <val> : interval is in hrs

        -timeranges "start=<time_stamp>,end=<time_stamp>,..." :

             Timestamp must be in format 'YYYY-MM-DD HH24:MI:SS'

    KPI for db:

        CPUPERCENT - CPU utilization - Percent

        IOREAD - Disk read - Mbyte/sec

        DBTIMEPERCALL - Database time per user call - usec/call

        IOWRITE - Disk write - Mbyte/sec

        IOTHROUGHPUT - Disk throughput - IO/sec

    KPI for cluster:

        CPUPERCENT - CPU utilization - Percent

        IOREAD - Disk read - Mbyte/sec

        IOWRITE - Disk write - Mbyte/sec

        IOTHROUGHPUT - Disk throughput - IO/sec

chactl remove model -name <model_name>

chactl rename model -from <model_name> -to <model_name>

chactl import model -name <model_name> -file <model_file> [-force]

chactl export model -name <model_name> -file <output_file>

chactl export repository -format mdb -start <time> -end <time>

chactl set maxretention -time <retention_time>

chactl resize repository -entities <total # of hosts and database instances> [-force | -eval]

chactl authorize user -name <db_user>

chactl revoke user -name <db_user>