Database Reference
In-Depth Information
GV$SERVICEMETRIC
This contains metric values measured for all the services defined in the database. The MMON process updates these
values as it captures load and other service-related information from the SGA. Updates to this view happen in
5-second and 1-minute intervals.
Apart from making updates to the GV$SERVICEMETRIC view, MMON also updates the operating system statistics and
uses this information to determine the load characteristics on the various nodes in the cluster. The following query
output from GV$OSSTAT provides the operating system statistics:
SELECT * FROM (SELECT INST_ID,STAT_NAME,VALUE FROM GV$OSSTAT)
PIVOT (SUM(VALUE) FOR INST_ID IN (1,2,3,4)) ORDER BY STAT_NAME
STAT_NAME 1 2 3 4
-------------------------- ---------- ---------- ---------- ----------
BUSY_TIME 68523311 137810899 69100616 63652298
FREE_MEMORY_BYTES 1.36E+08 1.06E+08
GLOBAL_RECEIVE_SIZE_MAX 4194304 4194304 4194304 4194304
GLOBAL_SEND_SIZE_MAX 4194304 4194304 4194304 4194304
IDLE_TIME 2263573618 2194213550 2848860288 2854392901
INACTIVE_MEMORY_BYTES 1.79E+09 1.72E+09
IOWAIT_TIME 15425632 40378746 10725915 6031263
LOAD 1.01953125 2.07910156 1.19921875 .639648438
NICE_TIME 236916 447162 932710 16292
NUM_CPUS 24 24 24 24
NUM_CPU_CORES 24 24 24 24
NUM_CPU_SOCKETS 4 4 4 4
PHYSICAL_MEMORY_BYTES 1.3527E+11 1.3527E+11 1.3527E+11 1.3527E+11
RSRC_MGR_CPU_WAIT_TIME 189 1550 185 131
SYS_TIME 10010888 19905107 12695839 10304209
TCP_RECEIVE_SIZE_DEFAULT 4194304 4194304 4194304 4194304
TCP_RECEIVE_SIZE_MAX 4194304 4194304 4194304 4194304
TCP_RECEIVE_SIZE_MIN 4096 4096 4096 4096
TCP_SEND_SIZE_DEFAULT 4194304 4194304 4194304 4194304
TCP_SEND_SIZE_MAX 4194304 4194304 4194304 4194304
TCP_SEND_SIZE_MIN 98304 98304 98304 98304
USER_TIME 54526555 105473193 51760395 51184651
VM_IN_BYTES 3.3496E+12 2.8286E+12 2.1828E+12 3.5692E+12
VM_OUT_BYTES 2.5269E+12 3.6358E+12 2.4633E+12 7.8788E+11
GV$OSSTAT
This contains O/S statistics updated by the MMON process and is used to determine the load on the nodes/servers.
The values are in hundredths of a second as a processor has been busy executing code and is averaged over
all processors.
 
Search WWH ::




Custom Search