HP XP Performance Advisor Software - Manual (Page 391)

Table 40 Metrics and descriptions (continued)
Description Metric Metric category
example, if collection
interval is set to five minutes,
the maximum port IO value
is collected and verified
against the previous value
every one minute, and the
highest port I/O value
collected in the last five
minutes is considered.
The average frontend I/Os
rate on this port over an Average Port IO – Frontend
entire collection interval. Port
Average IO shown in PA is
average of average IO.
Average Port IO is the value
within a period of sampling
time at one minute interval
from RMLIB call (this could
be running average in a
min). PA collects this data
every minute and updates
according to the configured
port collection frequency.
For example, consider port
collection is configured for
5 minutes, then every minute
average port data is
collected and stored
in-memory. At the end of 5th
minute, average of all
collected average port data
is calculated.
The minimum frontend I/Os
rate on this port over an Minimum Port IO – Frontend
entire collection interval. For
example, if collection
interval is set to five minutes,
the minimum port IO value
is collected and verified
against the previous value
every one minute, and the
lowest port I/O value
collected in the last five
minutes is considered.
The total IO is obtained by
summing up all the IO of all The total IO of all the LDEVs
created in a specified RAID Total IO – Frontend
the LDEVS for the given host
group. group for the given Host
group over the entire data
collection interval.
NOTE: If the same host
group is exposed to more
than one PORT then the IOs
to hostgroups is seen on all
PORTS where it is exposed.
Calculation of Total IO
depends on the LDEVS and
not on PORTS, while
summing up, the same
hostgroup is exposed to
more than one PORT, we
are considering only once.
Metrics and descriptions 391