The following shows the statistical information about the global buffer pool:

Reference note

The global buffer pool statistical information includes information about in-memory data buffers; however, the value output for each item is an approximate value.

 
pdstedit VV-RR(Object Option) ***** GLOBAL BUFFER INFORMATION *****
INPUT          :/tmp/stjdata [1]
OUTPUT RANGE   :**/**/** **:**:** - **/**/** **:**:** [2]
-----------------------------------------------------------------------------
HOST = test [3]
-----------------------------------------------------------------------------
EDIT TIME 2003/04/03 22:00:00 - 2003/04/03 23:00:00 [4]
               [5]   [6]   [7]   [8]  [9]    [10]  [11]  [12] [13]  [14]
         *LN1* SYNCW MAXB  UPGET UPHIT(HIT)  UPFLS RFGET RFHIT(HIT) RFFLS
               ----- ----- ----- ----- ---   ----- ----- ----- ---  -----
               [15]  [16]  [17]  [18]  [19]  [20]  [21]  [22] [23]  [24]
         *LN2* READ  WRITE WAITR WAITW WAITL BFINS PRRED PRHIT(HIT) PRINS
               ----- ----- ----- ----- ----- ----- ----- ----- ---  -----
               [25]  [26]  [27]  [28]  [29]  [30]  [31]  [32]       [33]
         *LN3* GBHIT CURRF CURUP TRGUP SYNCC PRRDR LRDRC LWTRC      LBBKR
               ----- ----- ----- ----- ----- ----- ----- -----      -----
               [34]  [35]  [36]  [37]  [38]     [39]  [40]     [41]  [42]
         *LN4* LBBKW CINSM CFMAX CFAVG SLEPC    SLEPR SLEPA    SPINR SPINA
               ----- ----- ----- ----- -------- ----- -------- ----- -----
               [43]  [44]  [45]  [46]  [51]  [52]
         *LN5* SYNCL SYNCB ALTRW ALTUW BUFWT BUFWQ
               ----- ----- ----- ----- ----- -----
SERVER :  sds [47]
  *BUFFER NAME:bp01 [48]            BUFFER:    100 [49]
         *LN1*     0     3 2.19k 1.55k( 37)    264 2.04k 1.42k( 33)   702
         *LN2*   663   692     0     6     0     0     0     0(  0)     0
         *LN3*    70 1.60k     0   800     4     0     0     0          0
         *LN4*     0     0     0  ****  0.0e+00   0.0  *******   0.8  277
         *LN5*  178k     0     0     0    57     0
  *BUFFER NAME:bp02                 BUFFER:    100
         *LN1*     0     4 2.19k 1.56k( 37)    302 2.04k 1.44k( 34)   640
         *LN2*   638   692     0     5     0     0     0     0(  0)     0
         *LN3*    71 1.60k     0   800     4     0     0     0          0
         *LN4*     0     0     0  ****  2.0e+00   0.0  1.5e+00   1.0  356
         *LN5*  178k     0     0     0 3.07k     0
  *BUFFER NAME:bp03                 BUFFER:    100
         *LN1*     0     3 2.61k 2.38k( 24)      0 7.35k 7.16k( 72)   184
         *LN2*   372 1.15k     0     1    10     0     0     0(  0)     0
         *LN3*    96 1.60k     0   800     8     0     0     0          0
         *LN4*     0     0     0  ****  5.0e+00   0.0  1.2e+00   0.5  481
         *LN5*  259k     0     0     0 8.70k     1
  *TOTAL* [50]
         *LN1*     0     4 7.00k 5.49k( 30)    566 11.4k 10.0k( 54) 1.53k
         *LN2* 1.67k 2.54k     0    12    10     0     0     0(  0)     0
         *LN3*    84 4.79k     0 2.40k    16     0     0     0          0
         *LN4*     0     0     0  ****  7.0e+00   0.0  1.3e+00   0.7  380
         *LN5*  205k           0     0 11.8k     1
-----------------------------------------------------------------------------
FILE KIND   LOG KIND   FIRST                 LAST                         NUM
STJ         buf        2012/02/03 09:30:00   2012/02/03 09:36:19           48
-----------------------------------------------------------------------------
 
NO FILE KIND:LOG FILE NAME
  LOG KIND             FIRST                 LAST                         NUM
 1 STJ      :/tmp/stjdata/pdstj01
  sys                  2012/02/03 09:30:49   2012/02/03 09:31:50            4
  uap                  2012/02/03 09:29:26   2012/02/03 09:31:04           26
  sql                  2012/02/03 09:29:25   2012/02/03 09:31:04          440
  sop                  2012/02/03 09:29:27   2012/02/03 09:31:04           98
  dop                  2012/02/03 09:29:27   2012/02/03 09:31:04           92
  pcd                  ****/**/** **:**:**   ****/**/** **:**:**            0
  obj                  ****/**/** **:**:**   ****/**/** **:**:**            0
  sqh                  ****/**/** **:**:**   ****/**/** **:**:**            0
  buf                  2012/02/03 09:30:00   2012/02/03 09:32:24           21
  fil                  2012/02/03 09:30:00   2012/02/03 09:32:24          224
  dfw                  2012/02/03 09:29:30   2012/02/03 09:32:24           71
  idx                  2012/02/03 09:30:00   2012/02/03 09:32:24            7
  fsv                  ****/**/** **:**:**   ****/**/** **:**:**            0
  hba                  ****/**/** **:**:**   ****/**/** **:**:**            0
 2 STJ      :/tmp/stjdata/pdstj02
  sys                  2012/02/03 09:34:58   2012/02/03 09:37:00            6
  uap                  2012/02/03 09:34:00   2012/02/03 09:36:17           19
  sql                  2012/02/03 09:33:59   2012/02/03 09:36:17          456
  sop                  2012/02/03 09:34:02   2012/02/03 09:36:16           65
  dop                  2012/02/03 09:34:02   2012/02/03 09:36:16           96
  pcd                  ****/**/** **:**:**   ****/**/** **:**:**            0
  obj                  ****/**/** **:**:**   ****/**/** **:**:**            0
  sqh                  ****/**/** **:**:**   ****/**/** **:**:**            0
  buf                  2012/02/03 09:34:10   2012/02/03 09:36:19           27
  fil                  2012/02/03 09:34:10   2012/02/03 09:36:19          288
  dfw                  2012/02/03 09:34:04   2012/02/03 09:36:19           41
  idx                  2012/02/03 09:34:10   2012/02/03 09:36:19            9
  fsv                  ****/**/** **:**:**   ****/**/** **:**:**            0
  hba                  ****/**/** **:**:**   ****/**/** **:**:**            0
 

Explanation

1. Name of input statistics unload file or name of the directory containing the input statistics unload file (maximum of 58 bytes)
2. Output range (output start date/time to output end date/time)
3. Name of host requesting output of global buffer pool statistical information
4. Edit period (collection start time to collection end time)
5. Number of pages output during synchronization point dump
   This is the maximum number of pages output from the global buffer pool to the HiRDB file during a synchronization point dump.
   If this value is large, many pages may be updated in the corresponding global buffer pool, and the updated page output rate may be low during a deferred write trigger, resulting in many pages kept in the global buffer pool. Evaluate whether the updated page output rate can be reduced during a deferred write trigger.
   
6. Maximum number of concurrently requested buffer sectors
   This is the number of global buffer pool acquisition requests issued concurrently from multiple transactions.
   If this value is almost the same as the number of global buffer sectors and the buffer shortages count is not 0, the buffer may be insufficient.
   
7. Update GETs count
   This is the number of times buffer acquisition was requested to the global buffer pool for updating purposes.
   
8. Update buffer hits count
   This is the number of times a page requested for the global buffer pool for updating purposes was found as the updated buffer.
   
9. Update buffer hit rate (%)
   This is the ratio of update buffer hits count to the total of update GETs count and reference GETs count. It is the probability of finding a requested page (for updating or referencing purposes) as updated buffer in the global buffer pool without executing a physical input/output operation.
   This value is obtained from the following formula:
   

   Update buffer hit rate = (update buffer hits count  (reference GETs count + update GETs count))  100

10. Update buffer flushes count
    This is the number of times the buffer was invalidated after writing its contents to a HiRDB file to create an empty buffer for reading new pages, because the referenced buffer had been updated.
    If this value is large, the deferred write processing may not be effective.
    If this is the case, either increase the number of global buffer pools or reduce the updated output page rate during deferred write trigger.
    
11. Reference GETs count
    This is the number of times buffer acquisition was requested for the global buffer pool for referencing purposes.
    
12. Reference buffer hits count
    This is the number of times a page requested for the global buffer pool for referencing purposes was found as the referenced (not updated) buffer.
    
13. Reference buffer hit rate (%)
    This is the ratio of reference buffer hits count to the total of reference GETs count and update GETs count. It is the probability of finding a requested page (for referencing or updating purposes) as referenced buffer in the global buffer pool without executing a physical input/output operation.
    This value is obtained from the following formula:
    

    Reference buffer hit rate = (reference buffer hits count  (reference GETs count + update GETs count))  100

14. Reference buffer flushes count
    This is the number of times the buffer was invalidated to create an empty buffer for reading new pages, because the referenced buffer had not been used since that time.
    
15. Actual READs count
    This is the number of times pages were read from HiRDB file.
    
16. Actual WRITEs count
    This is the number of times pages were written to a HiRDB file.
    
17. Input waits count
    This is the number of times wait status occurred because a requested page was being read from a HiRDB file by another user.
    
18. Output waits count
    This is the number of times wait status occurred because a requested page was being written to a HiRDB file by another user.
    
19. Buffer lock release waits count
    This is the number of times wait status occurred because a requested page was being used by another user.
    
20. Buffer shortages count
    This is the number of times there was no available buffer because all buffers (defined buffer sectors) were being used for referencing or update processing.
    This value must be 0.
    
21. Number of prefetched input pages
    This is the number of pages read in memory as being subject to prefetch processing.
    
22. Prefetch hits count
    This is the hits count on the pages that were read in memory as being subject to prefetch processing.
    
23. Prefetch hit rate (%)
    This is the ratio of the prefetch hits count to the number of prefetched input pages.
    
24. Prefetch resource shortages count
    This is the number of times a shortage occurred on the resources required for prefetch processing.
    
25. Global buffer hit rate (%)
    The guideline for the global buffer hit rate is 80%. If this value is 80% or less, you need to increase the number of buffer sectors.
    
26. Number of current reference buffers
27. Number of current update buffers
28. Number of update buffers that triggers output operation when deferred write trigger is on
29. Number of synchronization point dumps that occurred
30. Number of prefetch READ requests
31. Number of LOB global buffer READ requests
32. Number of LOB global buffer WRITE requests
33. Number of LOB global buffer pages input in batch mode
34. Number of LOB global buffer pages output in batch mode
35. Information used by the system (not for the user)
36. Information used by the system (not for the user)
37. Information used by the system (not for the user)
38. Number of contention-lock release waits resulting from lock processing on the global buffer pool (the entire buffer defined in one pdbuffer)
39. Percentage of contention-lock release waits resulting from lock processing on the global buffer pool (the entire buffer defined in one pdbuffer) (%)
40. Average sleep processing execution count of lock processing on the global buffer pool (the entire buffer defined in one pdbuffer)
41. Percentage of lock processing in spin processing on the global buffer pool (the entire buffer defined in one pdbuffer) (%)
42. Average spin loop execution count during lock processing on the global buffer pool (the entire buffer defined in one pdbuffer)
43. Buffer pool lock exclusive time during synchronization point processing (microseconds)
44. Number of buffers processed within buffer pool lock exclusive time during synchronization point processing
45. Take-over count of database write processing by reference request hit during synchronization point processing
46. Take-over count of database write processing by update request hit during synchronization point processing
47. Server name
    This is the server requesting output of global buffer pool statistical information.
    
48. Global buffer pool name
    This is the name of the global buffer pool for which statistical information was edited.
    
49. Number of buffer sectors in the global buffer pool
    This is the number of buffer sectors in the corresponding global buffer pool.
    
50. Totals for each item
51. Contention-lock release wait time during global buffer lock acquisition processing (microseconds)^{#1, #2}
52. Number of requests in the contention-lock release queue during global buffer lock acquisition processing

#1

Because the precision of the displayed values depends on the OS and the hardware being used, a displayed value might be smaller than the actual value.

#2

This information is output when the -w option is specified in the pdstbegin command to collect the thread-to-thread lock-release wait time. For details about the -w option in the pdstbegin command, see (4) Options under 2.109 pdstbegin (Start output of statistical information).