|
|
|
Source: ASTAVO file at the DAYS timespan
HARDVA file at the DAYS timespan
Function: This facility allows the storage administrator to
monitor key storage indicators related to the
operation of DASD cache control units. While DASD
cache has become an important part of the total
storage hierarchy, the lack of convenient and
powerful reporting tools has made it difficult to
manage this resource. The introduction of ASTEX
data into CA MICS should help correct this
problem, because it gives storage administrators
the ability to quickly and easily monitor this
important resource.
Cache-related problems are normally caused by too
many users needing cache resources at the same
time. Detection and correction of these problems
requires the ability to analyze the data are
varying levels of granularity. This facility
allows selection and summarization by specific
dates, hours, and CA MICS zones. Selection can
also be limited to the data for a specific volume,
SMS storage group, ASTEX contention group, or MVS
logical control unit. This type of multi-
dimensional analysis can detect problems that
might otherwise remain hidden.
Features: As stated above, a wide variety of options are
available in terms of data selection and
summarization. These options make it easy to run
several reports at varying levels of detail to
quickly identify and isolate a problem.
Output reports from this facility include tabular
reports, printer plots, or color graphic plots.
When plots are requested, up to nine different
elements may be plotted on the same chart.
SMS
Issues: The ability to select and summarize by an SMS
storage group allows you to monitor and tune your
cache resources and your SMS groups so they
deliver optimum performance. The ability to
summarize the report using a Volume Group Table
allows you to simulate the operation of alternate
storage group definitions. Thus, you can easily
move volumes between storage groups, monitor the
results, and then finally make the change in your
group definitions when the results are acceptable.
ELEMENT DESCRIPTIONS
SYSID/S
This value appears at the top of each page and
identifies the System Identifier (system) to which
this data applies. The shorter value is used with
color graphic plots.
Volume/Vol
Storage Group/Grp
Logical Ctl Unit/LCUID
Contention Group/CGrp
The top line of each page also indicates the
summary group to which the values on this plot
apply. Depending on the report option chosen,
this summarization will be done by volume, SMS
storage group, MVS logical control unit, or ASTEX
contention group. The shorter value listed above
is used with color graphic plots.
Date
When summarization by hour or zone is requested,
the date will appear at the top of each chart to
indicate the date for which the plotted data
applies.
Vertical Axis (Value)
The numeric values distributed vertically down the
left side of each page represent the scale of the
numeric values being plotted on that page. When
requesting color graphics plots, you can specify
the values that will appear on this axis. This
should be done with care, because all plots and
groups will have to adapt to the values you
specify. For printer plots, or when values are
not specified, SAS will generate the values
automatically, based on the content of the plotted
data. We suggest that these SAS defaults be used
whenever possible, because they will avoid many
potential problems. Care should also be taken
when specifying multiple elements on one plot,
that the elements chosen will have a similar range
of values.
Horizontal Axis (Date/Zone/Hour)
Plots on the horizontal axis represent the date,
hour, or zone corresponding to the value being
plotted. Report options are used to determine
which of these three values will be used. When
date summarization is specified, the date appears
in the format 'ddmmmyy', and the number of dates
plotted is determined by the number of cycles of
CA MICS data specified as input to the report.
Footnotes
The footnote(s) section of each graph is
important, as it indicates the names of the data
element(s) that are plotted on the graph. Up to 9
elements can be plotted on the same graph, in
which case 9 footnotes will appear at the bottom
of each page. For printer plots, an alphabetic
character A-I will appear to the left of each
description, identifying the character used on the
plot to represent that particular element. On
color graphic plots, the letter does not appear,
because the color of the footnote should
correspond with the color of the plot point/line
on the graph. If different colors do not appear
for each point and footnote, your MICF options
should be modified to specify different colors. A
brief description of the possible values in the
footnote section, and what those values represent,
is given below:
I/O Requests: Total
The total number of I/O requests received for
this group during this measurement interval.
I/O Requests: Cache Candidate
The number of I/O requests received that were
considered to be eligible for caching.
I/O Requests: Read The total number of read I/O
requests that were considered to be eligible
for caching.
I/O Requests: Read Sequential
The total number of read sequential I/O
requests that were considered to be eligible
for caching.
I/O Requests: DASD Fast Write
The number of I/O requests received that were
considered candidates for DASD Fast Write
caching.
I/O Requests: Cache Fast Write
The Number of I/O requests received that were
considered candidates for Cache Fast Write
caching.
I/O Requests: Non-Cache Candidate
The total number of I/O requests received that
were not considered to be cache candidates.
I/O Requests: Write
The total number of I/O requests received that
were not considered to be cache candidates
because they were write requests.
I/O Requests: Bypass Cache
The total number of I/O requests received that
were not considered to be cache candidates
because they requested no caching.
I/O Requests: IOS Inhibited
The total number of I/O requests received that
were not considered to be cache candidates
because IOS had inhibited them due to cache
activity.
I/O Requests: ASTEX Inhibited
The total number of I/O requests received that
were not considered to be cache candidates
because ASTEX had inhibited them due to cache
activity.
I/O Requests: Other
The total number of I/O requests received that
were not considered to be cache candidates for
reasons other than those listed above.
Cache Effective: Total
The total number of I/O requests that were
successfully processed by cache (cache hits).
Cache Effective: Read
The total number of read I/O requests that
were successfully processed by cache (cache
hits).
Cache Effective: Read Sequential
The total number of read sequential I/O
requests that were successfully processed by
cache (cache hits).
Cache Effective: DASD Fast Write
The total number of I/O requests that were
successfully processed by DASD Fast Write
cache.
Cache Effective: Cache Fast Write
The total number of I/O requests that were
successfully processed by Cache Fast Write
cache.
Total Tracks Loaded into Cache
The total number of tracks loaded into cache
during this measurement interval.
Total DASD Fast Write Track Loads
The total number of tracks loaded into DASD
Fast Write cache during this measurement
interval.
NVS Overflow Percentage
The percentage of times that requests for
non-volatile storage exceeded NVS capacity and
caused an overflow condition.
Average Demotion Time
The average amount of time a storage page
stayed in cache memory, before being replaced
by a more active page.
Percent Effective: All I/O
The average percentage of all I/O requests
received that were handled by cache storage.
Percent Effective: Cache Candidates
The average percentage of all cache candidate
I/O requests received that were handled by
cache storage.
Percent Effective: Read
The average percentage of all read I/O
requests received that were handled by cache
storage.
Percent Effective: Read Sequential
The average percentage of all read sequential
I/O requests received that were handled by
cache storage.
Percent Effective: DASD Fast Write
The average percentage of all DASD Fast Write
requests received that were handled by DASD
Fast Write.
Percent Effective: Cache Fast Write
The average percentage of all Cache Fast Write
requests received that were handled by Cache
Fast Write.
Percent Demotion Time < 30 seconds
The percentage of time during this interval
that the average demotion time was less than
30 seconds. Demotion time indicates the
amount of time a page can remain unreferenced
before being removed from cache.
Percent Demotion Time < 1 minute
The percentage of time during this interval
that the average demotion time was less than 1
minute. Demotion time indicates the amount of
time a page can remain unreferenced before
being removed from cache.
Percent Demotion Time < 2 minutes
The percentage of time during this interval
that the average demotion time was less than
two minutes. Demotion time indicates the
amount of time a page can remain unreferenced
before being removed from cache.
Percent Demotion Time > 2 minutes
The percentage of time during this interval
that the average demotion time was greater
than two minutes. Demotion time indicates the
amount of time a page can remain unreferenced
before being removed from cache.
DAILY ASTEX CACHE STATISTICS (STGEPE) CA MICS StorageMate ABC CORPORATION SYSID=SYSA Storage Group=TSOLRG Date=05APR91 175000 + | | | A | B 150000 + | | | | 125000 + V | A a | B l | u | A e 100000 + | A | A A | B | 75000 + | B A | B | A B | A B 50000 + A B | B A | B | A B | 25000 + A B | B A A A | B B B A | A A B A | B B A A A B A 0 + B B B B ----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+----+ 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Hour A = I/O Requests: Cache Candidate B = Cache Effective: Total
Figure 4-20. Daily ASTEX Cache Statistics / Plot
| Copyright © 2012 CA. All rights reserved. | Tell Technical Publications how we can improve this information |