Monitoring Directory Search Activity with PMOMON › D7: Private Library Searches – Summary

D7: Private Library Searches – Summary

The D7 display summarizes private library directory searches, as opposed to BLDL/FIND requests. See D2: BLDL/FIND Request Activity in this chapter for information on the difference between library directory searches and BLDL/FIND requests.

The following screen illustrates a sample D7 display:

PMO 4.4 ( Active )  Sys=IPOX  13NOV10:09:10:50  PMO Active Time=  5.500 hrs

                D7   Private Library Searches - Summary

                                                 % of   Avg Elap   Elap Time
                                        Count    Total     Time     Savings

  Private Library Searches         = 1,951,145   100   0.000846S    1.24D

      Hash Tables                  = 1,925,062    99   0.000100S    1.24D
        Member Found               =   107,059     5                1.65H
        Member Not Found           = 1,818,003    93                1.17D

     I/O Library Searches          =   26,083      1   0.055892S
         Updates/Concatenated Lib  =    8,060      0
           Member Found            =        0      0
           Member Not Found        =    8,060      0
         Hash Table Candidates     =   18,023      1
           Member Found            =    5,922      0
           Member Not Found        =   12,101      1

Row Headings

The row headings on the D7 display are described in the following:

Private Library Searches

Directory searches of private libraries.

Hash Tables

Directory searches of private libraries that the hash tables resolved.

Member Found

Directory searches of private libraries in which the member was found.

Member Not Found

Searches in which the member was not found.

When CA PMO cannot find the sought-for member in a hash table, this member does not exist. An I/O search is not needed to determine that the member cannot be found.

Note: Hash table searches and I/O searches can always determine that a member is not in a private library.

I/O Library Searches

Directory searches resolved by DASD I/O.

Updates/Concatenated Lib

For a small percentage of searches, libraries with hash tables will require DASD I/O. There are two reasons for the DASD I/O:

If CA PMO is updating the hash table during a search, the search is turned over to DASD I/O.

If a library with a hash table is earlier in a concatenation than a library without a hash table, and if the sought-for member is not in the hash table, an I/O directory search takes place against all libraries in the concatenation up to the library that contains the sought-for member (or until the directories of all of the libraries in the concatenation have been searched, if the member is not in any of the libraries).

Note: To determine whether I/O searches are due to updates or concatenation problems, you can produce an offline report for the library. For details, see the chapter "Producing Offline Reports for Managed Private Libraries."

The D8 display shows the names of libraries with hash tables that are having I/O searches due to concatenation problems. In general, you can solve this problem by allowing CA PMO to manage all private libraries automatically. If you absolutely must rely on manual selection of the private libraries to be managed, see the example in the section Tuning CA PMO After Installation in the chapter "Tuning CA PMO" for information on this problem and how to solve it.

For I/O searches due to concatenation problems, only the count for Member Not Found is incremented in this field. This is because a concatenation problem occurs only when the sought-for member resides in a library that does not have a hash table, and this field describes only those libraries that do have hash tables.

Hash Table Candidates

I/O searches for members of private libraries that do not have hash tables. This includes any private library that CA PMO is not managing. The D9 display shows a list of candidate libraries.

Column Headings

The column headings on the D7 display are described in the following:

Count

The number of private library directory searches. For a discussion of how CA PMO categorizes and counts library searches, see the discussion of the Count field earlier in this chapter.

% Of Total

A breakdown of library directory searches by type of search.

Avg Elap Time

The average time to resolve a search for a private library member.

For hash table searches, CA PMO calculates average elapsed time when the searches are actually resolved. If search times for hash tables approach those for DASD I/O, you need to tune CA PMO. For details, see the chapter "Tuning CA PMO," the sections Performing Initial Tuning of CA PMO and Additional Tuning at Installation Time.

For I/O searches, CA PMO calculates average elapsed time during a directory search sampling period. CA PMO samples I/O directory search times just before it actually begins managing a library. This prevents CA PMO from influencing average elapsed times for I/O library searches.

CA PMO may use a simple average or a weighted average for I/O searches:

• If there were no I/O searches since CA PMO became active or since the last F PMO,RESET command, CA PMO averages the search times that it collected during the sample searches.
• If there were I/O searches, CA PMO uses a weighted average, based on the simple average that it calculated from the sample searches. CA PMO weights the simple average according to the number of DASD searches for a particular library while CA PMO is active.

Elap Time Savings

The amount of search time that CA PMO saved, based on the difference between averaged elapsed time for I/O and times for either the managed list or the hash tables (as appropriate).

If you see three asterisks (***), CA PMO did not save any elapsed time.