|
 |
|
Monitor Data Guard I/O Activity using iostat
Donald K. Burleson
|
Also see these important notes on
Data
Guard redo transport waits
The performance of the log transfer service to
a standby site depends on the I/O subsystem and the speed of I/O
activity on the disk where archived redo logs are being written. To
determine if the I/O on the standby host is the bottleneck in the
log transfer process, the iostat operating system utility, or
a variant of it depending on the operating system, can be used. The
iostat utility shows a detailed map of the I/O activity on
the disks. The important columns for log transfer service are:
-
Kw/S (Average KB written per
second)
-
Wait
-
%w (Occupancy of wait queue)
-
%b (device busy)
-
asvc_t (average service time)
If there is a lot of I/O wait in the system,
the disks should be configured properly, using RAID or any other
technology, to boost the efficiency of I/O activity.
The following is sample output from iostat –xn
5 30 that shows I/O statistics collected every 5 seconds for 30
intervals. Note a %w and asvc_t for the device c0t0d0. This disk
contains the log archive destination for the physical standby
database “appsdb”. This I/O wait will result in a delay in
completion of the transfer of archived log files to the standby
host.
DBA@jrsp01 iostat -xn 5 20
I/O Activity on the Standby Host
r/s w/s kr/s kw/s
wait actv wsvc_t asvc_t %w %b device
56.6 382.8 1079.1 3991.9 3.3 1.9 7.6 4.3 81 97 c0t0d0
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0 0 c0t2d0
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0 0 c0t3d0
|
