HP PolyServe Software - Manual (Page 9)

Advanced I/O Monitoring 6
MxODM I/O Monitoring: Practical Examples
This section offers several practical examples of the powerful I/O monitoring
capability of the mxodmstat(8) command. The examples use a question and
answer format, where the answer includes an mxodmstat(8) command and its
output.
The system used for these examples was a six-node, dual-processor, Intel-based
cluster with PolyServe Matrix Server 2.1 and Oracle9i Real Application Clusters
version 9.2.0.4. There were two main databases on the cluster:
• BENCH. This database is a stress-testing environment used at PolyServe. It is
an OLTP workload that exhibits realworld characteristics. The BENCH database
has four instances executing on nodes 1 through 4.
• DEV. This database contains a small application under development that
simulates capturing stock trading events. For this set of examples, there were
two instances accessing the
DEV database. The DEV1 instance was executing on
node 5 of the cluster and the DEV2 instance was on node 6.
Having two databases, one with four instances and another with two instances, is
a good case for demonstrating the invaluable information that DBAs can obtain
with the mxodmstat(8) command.
Example 1
Question: What is the cluster-wide read and write count, I/O throughput (Kbytes),
and average I/O response times with 3-second granularity? How about the read-
to-write ratio?
Answer: This information can be monitored with the -a and -p options of
mxodmstat(8). The op argument directs mxodmstat(8) to break out reads and
writes. When the -p option is supplied, the dataset is enhanced to show I/O
percentages. The I/O service times are also displayed in Millisecond units. Figure
1 shows an example.