Exadata X2-2 1/4 RACK上的并行测试

Question

imageinfo
Kernel version: 2.6.18-274.18.1.0.1.el5 #1 SMP Thu Feb 9 19:07:16 EST 2012 x86_64
Image version: 11.2.3.1.1.120607
Image activated: 2012-08-14 19:16:01 -0400
Image status: success
System partition on device: /dev/mapper/VGExaDb-LVDbSys1
SQL> select * from v$version;
BANNER
Oracle Database 11g Enterprise Edition Release 11.2.0.2.0 - 64bit Production
PL/SQL Release 11.2.0.2.0 - Production
CORE    11.2.0.2.0      Production
TNS for Linux: Version 11.2.0.2.0 - Production
NLSRTL Version 11.2.0.2.0 - Production
SQL> alter system flush buffer_cache;
System altered.
SQL> //
System altered.
SQL>  set timing on;
SQL> select a.name,b.value
  2        from v$sysstat a , v$mystat b
      where
a.statistic#=b.statistic#
and (a.name in ('physical read total bytes','physical write total bytes',
'cell IO uncompressed bytes') or a.name like 'cell phy%'
or a.name like '%flash cache read hits');    3    4    5    6    7 
NAME                                                                  VALUE
physical read total bytes                                            114688
physical write total bytes                                                0
cell physical IO interconnect bytes                                  114688
cell physical IO bytes pushed back due to excessive CPU on cell           0
cell physical IO bytes saved during optimized file creation               0
cell physical IO bytes saved during optimized RMAN file restore           0
cell physical IO bytes eligible for predicate offload                     0
cell physical IO bytes saved by storage index                             0
cell physical IO interconnect bytes returned by smart scan                0
cell IO uncompressed bytes                                                0
cell flash cache read hits                                                0
11 rows selected.
Elapsed: 00:00:00.01
SQL> select count(*) from larget;
  COUNT(*)
242778112
Elapsed: 00:00:16.83
SQL>  set timing on;
SQL> select a.name,b.value
  2        from v$sysstat a , v$mystat b
      where
a.statistic#=b.statistic#
and (a.name in ('physical read total bytes','physical write total bytes',
'cell IO uncompressed bytes') or a.name like 'cell phy%'
or a.name like '%flash cache read hits');    3    4    5    6    7 
NAME                                                                  VALUE
physical read total bytes                                        2.6262E+10
physical write total bytes                                                0
cell physical IO interconnect bytes                              3018259592
cell physical IO bytes pushed back due to excessive CPU on cell           0
cell physical IO bytes saved during optimized file creation               0
cell physical IO bytes saved during optimized RMAN file restore           0
cell physical IO bytes eligible for predicate offload            2.6262E+10
cell physical IO bytes saved by storage index                             0
cell physical IO interconnect bytes returned by smart scan       3018112136
cell IO uncompressed bytes                                       2.6285E+10
cell flash cache read hits                                            18639
11 rows selected.
Elapsed: 00:00:00.01
SQL> select count(*) from larget;
  COUNT(*)
242778112
SQL> select /*+ parallel */  count(*) from larget;
  COUNT(*)
242778112
Elapsed: 00:00:02.71
SQL>  set timing on;                   
SQL> select a.name,b.value
  2        from v$sysstat a , v$mystat b
      where
a.statistic#=b.statistic#
and (a.name in ('physical read total bytes','physical write total bytes',
'cell IO uncompressed bytes') or a.name like 'cell phy%'
or a.name like '%flash cache read hits');    3    4    5    6    7 
NAME                                                                  VALUE
physical read total bytes                                        7.8787E+10
physical write total bytes                                                0
cell physical IO interconnect bytes                              9054570192
cell physical IO bytes pushed back due to excessive CPU on cell           0
cell physical IO bytes saved during optimized file creation               0
cell physical IO bytes saved during optimized RMAN file restore           0
cell physical IO bytes eligible for predicate offload            7.8787E+10
cell physical IO bytes saved by storage index                             0
cell physical IO interconnect bytes returned by smart scan       9054340816
cell IO uncompressed bytes                                       7.8855E+10
cell flash cache read hits                                            73375
11 rows selected.
Elapsed: 00:00:00.00
SQL> select /*+ parallel(larget 128) */  count(*) from larget;
  COUNT(*)
242778112
Elapsed: 00:00:09.02
SQL> SQL> select /*+ parallel(larget 64) */ count(*) from larget;
  COUNT(*)
242778112
Elapsed: 00:00:02.33通过_px_trace了解 并行详情：
SQL> oradebug setmypid;
Statement processed.
SQL> oradebug tracefile_name;
/u01/app/oracle/diag/rdbms/dbm/dbm1/trace/dbm1_ora_10230.trc
SQL> alter session set "_px_trace"="compilation","execution","messaging";
Session altered.
Elapsed: 00:00:00.00
SQL> select /*+ parallel */  count(*) from larget;
  COUNT(*)
242778112
Elapsed: 00:00:02.63
kxfpsori
        Sorted: 1(2400:1085669688) 2(2400:1085669760)
kxfplist
        Getting instance info for default group
2012-09-02 10:28:29.527933*:PX_Messaging:kxfp.c@22076:kxfpGetTotalCpuCount(): kxfplist returned status: 2
2012-09-02 10:28:29.527933*:PX_Messaging:kxfp.c@17529:kxfpiinfo():      inst [cpus:mxslv]
2012-09-02 10:28:29.527933*:PX_Messaging:kxfp.c@17533:kxfpiinfo():      1    [24  :128  ]
2012-09-02 10:28:29.527933*:PX_Messaging:kxfp.c@17533:kxfpiinfo():      2    [24  :128  ]
2012-09-02 10:28:29.527933*:PX_Messaging:kxfp.c@22112:kxfpGetTotalCpuCount(): instance_id: 1, cpu_count: 24
2012-09-02 10:28:29.527933*:PX_Messaging:kxfp.c@22112:kxfpGetTotalCpuCount(): instance_id: 2, cpu_count: 24
2012-09-02 10:28:29.527933*:PX_Messaging:kxfp.c@18160:kxfpunit():       var=0 limit=768 use_aff=0 aff_num=0 unit=24
kxfpgsg
        getting 1 sets of 12 threads, client parallel query execution flg=0x230
        Height=0, Affinity List Size=0, inst_total=2, coord=1
        Insts     1  2
        Threads   0 12
kxfpg1sg
        q:0x457f2e610 req_threads:12 nthreads:12 unit:24 #inst:2 normal
        jStart:0  jEnd:128  jIncr:1 isGV:0  i:0 instno:1 kxfpilthno:0
        jStart:0  jEnd:128  jIncr:1 isGV:0  i:1 instno:2 kxfpilthno:12
kxfpg1srv
        trying to get slave P000 on instance 2 for q:0x457f2e610
        slave P000 is remote (inst:2)
        Slave P000 acquired dp:(nil)
        Got It. 1 so far.
kxfpg1srv
        trying to get slave P001 on instance 2 for q:0x457f2e610
        slave P001 is remote (inst:2)
        Slave P001 acquired dp:(nil)
        Got It. 2 so far.
kxfpg1srv
        trying to get slave P002 on instance 2 for q:0x457f2e610
        slave P002 is remote (inst:2)
kxfpg1sg
        got 12 servers (sync), errors=0x0 returning
2012-09-02 10:28:29.532850*:PX_Messaging:kxfp.c@10138:kxfpgsg():        Acquired 12 slaves on 1 instances avg height:12 #set:1 qser:2835457
2012-09-02 10:28:29.532850*:PX_Messaging:kxfp.c@10148:kxfpgsg():        P000 inst 2 spid 10369
2012-09-02 10:28:29.532850*:PX_Messaging:kxfp.c@10148:kxfpgsg():        P001 inst 2 spid 10372
2012-09-02 10:28:29.532850*:PX_Messaging:kxfp.c@10148:kxfpgsg():        P002 inst 2 spid 10374
2012-09-02 10:28:29.532850*:PX_Messaging:kxfp.c@10148:kxfpgsg():        P003 inst 2 spid 10376
2012-09-02 10:28:29.532850*:PX_Messaging:kxfp.c@10148:kxfpgsg():        P004 inst 2 spid 10378
2012-09-02 10:28:29.532850*:PX_Messaging:kxfp.c@10148:kxfpgsg():        P005 inst 2 spid 10380
2012-09-02 10:28:29.532850*:PX_Messaging:kxfp.c@10148:kxfpgsg():        P006 inst 2 spid 10382
2012-09-02 10:28:29.532850*:PX_Messaging:kxfp.c@10148:kxfpgsg():        P007 inst 2 spid 10384
2012-09-02 10:28:29.532850*:PX_Messaging:kxfp.c@10148:kxfpgsg():        P008 inst 2 spid 10386
2012-09-02 10:28:29.532850*:PX_Messaging:kxfp.c@10148:kxfpgsg():        P009 inst 2 spid 10388
2012-09-02 10:28:29.532850*:PX_Messaging:kxfp.c@10148:kxfpgsg():        P010 inst 2 spid 10393
2012-09-02 10:28:29.532850*:PX_Messaging:kxfp.c@10148:kxfpgsg():        P011 inst 2 spid 10395
2012-09-02 10:28:29.532850*:PX_Messaging:kxfp.c@10164:kxfpgsg():        Instance(servers):
2012-09-02 10:28:29.532850*:PX_Messaging:kxfp.c@10167:kxfpgsg():        inst:1 #slvs:0
2012-09-02 10:28:29.532850*:PX_Messaging:kxfp.c@10167:kxfpgsg():        inst:2 #slvs:12
kxfpValidateSlaveGroup
        qcq:0x457f2e610 flg:30
kxfxcp1
        Sending parse to nprocs:12 slave_set:1
kxfxcPutSession获得 1 set of 12 threads ,并行度为12， 这里可以看到Oracle Exadata自动的并行度还是比较合理的
如何在OTN中文技术论坛提一个问题？
论坛礼仪需知及学习oracle的方法论
Maclean Liu
Oracle Database Administrator
Oracle Certified 10g/11g Master     
www.askmaclean.com

Answer 1

忘记列出并行参数了：
Oracle Database 11g Enterprise Edition Release 11.2.0.2.0 - 64bit Production
With the Partitioning, Real Application Clusters, Automatic Storage Management, OLAP,
Data Mining and Real Application Testing options
SQL> show parameter parallel
NAME                                 TYPE        VALUE
fast_start_parallel_rollback         string      LOW
parallel_adaptive_multi_user         boolean     FALSE
parallel_automatic_tuning            boolean     FALSE
parallel_degree_limit                string      CPU
parallel_degree_policy               string      MANUAL
parallel_execution_message_size      integer     16384
parallel_force_local                 boolean     FALSE
parallel_instance_group              string
parallel_io_cap_enabled              boolean     FALSE
parallel_max_servers                 integer     128
parallel_min_percent                 integer     0
parallel_min_servers                 integer     32
parallel_min_time_threshold          string      AUTO
parallel_server                      boolean     TRUE
parallel_server_instances            integer     2
parallel_servers_target              integer     384
parallel_threads_per_cpu             integer     2
recovery_parallelism                 integer     0