绝对干货！|【性能调优】Oracle AWR报告指标全解析

数据与人 2020-12-15

25420

绝对干货！|【性能调优】Oracle AWR报告指标全解析

啥是AWR?

==============================

(Automatic Workload Repository)

一堆历史性能数据，放在SYSAUX表空间上， AWR和SYSAUX都是10g出现的，是Oracle调优的关键特性；
默认快照间隔1小时，10g保存7天、11g保存8天;

AWR程序核心是dbms_workload_repository包

@?/rdbms/admin/awrrpt 本实例
@?/rdbms/admin/awrrpti RAC中选择实例号

谁维护AWR?

主要是MMON(Manageability Monitor Process)和它的小工进程(m00x)
MMON的功能包括:
1.启动slave进程m00x去做AWR快照
2.当某个度量阀值被超过时发出alert告警
3.为最近改变过的SQL对象捕获指标信息

AWR小技巧
手动执行一个快照：

Exec dbms_workload_repository.create_snapshot;

创建一个AWR基线

Exec DBMS_WORKLOAD_REPOSITORY.CREATE_BASELINE(start_snap_id，end_snap_id ,baseline_name);
@?/rdbms/admin/awrddrpt AWR比对报告
@?/rdbms/admin/awrgrpt RAC 全局AWR

1、报告总结

WORKLOAD REPOSITORY report for

DB Name	DB Id	Instance	Inst num	Startup Time	Release	RAC
CUFS	3961207481	cufs	1	24-Aug-19 21:08	11.2.0.4.0	NO

Host Name	Platform	CPUs	Cores	Sockets	Memory (GB)
database	Linux x86 64-bit	16	16	4	31.48

	Snap Id	Snap Time	Sessions	Cursors/Session
Begin Snap:	9512	04-Dec-19 10:00:20	85	9.3
End Snap:	9513	04-Dec-19 11:00:28	80	3.4
Elapsed:		60.14 (mins)
DB Time:		64.77 (mins)

Elapsed为该AWR性能报告的时间跨度(自然时间的跨度，例如前一个快照snapshot是9点生成的，后一个快照snapshot是11点生成的，

则若使用@?/rdbms/admin/awrrpt 脚本中指定这2个快照的话，那么其elapsed = (11-9)=2 个小时)

一个AWR性能报告至少需要2个AWR snapshot性能快照才能生成 ( 这2个快照时间实例不能重启过，否则指定这2个快照生成AWR性能报告会报错)，

DB TIME= 所有前台session花费在database调用上的总和时间：

注意是前台进程foreground sessions
包括CPU时间、IO Time、和其他一系列非空闲等待时间

DB TIME 不等于响应时间，DB TIME高了未必响应慢，DB TIME低了未必响应快

DB Time描绘了数据库总体负载，但要和elapsed time结合其来。

Average Active Session AAS= DB time/Elapsed Time
DB Time =60min，Elapsed Time =60min AAS=60/60=1 负载一般
DB Time= 1min，Elapsed Time= 60min AAS= 1/60 负载很轻
DB Time= 60000min，Elapsed Time=60min AAS=1000 系统hang住
DB TIME= DB CPU + Non-Idle Wait + Wait on CPU queue

如果仅有2个逻辑CPU，而2个session在60分钟都没等待事件，一直跑在CPU上，那么：

DB CPU=2*60mins ， DB Time =2*60 +0 +0 =120
AAS = 120/60=2 正好等于OS load 2。
如果有3个session都100%仅消耗CPU，那么总有一个要wait on queue
DB CPU =2* 60mins ，wait on CPU queue=60mins
AAS= (120+ 60)/60=3 主机load 亦为3，

此时vmstat 看waiting for run time

1-1 内存参数大小

Cache Sizes

	Begin	End
Buffer Cache:	2,880M	2,880M	Std Block Size:	8K
Shared Pool Size:	4,782M	4,796M	Log Buffer:	4,356K

1-2 Load Profile

Load Profile

	Per Second	Per Transaction	Per Exec	Per Call
DB Time(s):	1.1	0.0	0.00	0.00
DB CPU(s):	1.1	0.0	0.00	0.00
Redo size (bytes):	35,814.6	140.3
Logical read (blocks):	22,708.3	88.9
Block changes:	230.3	0.9
Physical read (blocks):	395.7	1.6
Physical write (blocks):	4.6	0.0
Read IO requests:	8.7	0.0
Write IO requests:	2.3	0.0
Read IO (MB):	3.1	0.0
Write IO (MB):	0.0	0.0
User calls:	2,795.2	11.0
Parses (SQL):	635.8	2.5
Hard parses (SQL):	20.9	0.1
SQL Work Area (MB):	2.7	0.0
Logons:	0.1	0.0
Executes (SQL):	641.8	2.5
Rollbacks:	252.3	1.0
Transactions:	255.3

指标含义

redo size单位bytes，redosize可以用来估量update/insert/delete的频率，大的redo size往往对lgwr写日志，和arch归档造成I/O压力，

Per Transaction可以用来分辨是大量小事务,还是少量大事务。如上例每秒redo 约1MB,每个事务800 字节，符合OLTP特征

Logical Read 单位(次数*块数),逻辑读耗CPU,主频和CPU核数都很重要，逻辑读高则DB CPU往往高，也往往可以看到latch: cache buffer chains等待。

Block changes 单位(次数*块数)，描绘数据变化频率

Physical Read 单位(次数*块数)，物理读消耗IO读，体现在IOPS和吞吐量等不同纬度上；但减少物理读可能意味着消耗更多CPU。
好的存储每秒物理读能力达到几GB，例如Exadata。这个physical read包含了physical reads cache和physical reads direct

Physical writes 单位(次数*块数)，主要是DBWR写datafile，也有direct path write。dbwr长期写出慢会导致定期log file switch(checkpoint no complete) 检查点无法完成的前台等待。这个physical write 包含了physical writes direct +physical writes from cache

User Calls 单位次数，用户调用数，more details from internal

Parses 解析次数，包括软解析+硬解析，软解析优化得不好，则夸张地说几乎等于每秒SQL执行次数。即执行解析比1:1，而我们希望的是解析一次到处运行

Hard Parses 万恶之源:Cursor pin s on X， library cache: mutex X ， latch: row cache objects shared pool……………..。硬解析最好少于每秒20次

W/A MB processed 单位MB W/A workarea workarea中处理的数据数量，结合 In-memory Sort%

Logons 登陆次数，结合AUDIT审计数据一起看。

Executes 执行次数，反应执行频率

Rollback 回滚次数，反应回滚频率,但是这个指标不太精确，

Transactions 每秒事务数，是数据库层的TPS，可以看做压力测试或比对性能时的一个指标，孤立看无意义

% Blocks changed per Read 每次逻辑读导致数据块变化的比率；如果’redo size’, ‘block changes’ ‘pct of blocks changed per read’三个指标都很高，则说明系统正执行大量insert/update/delete;pct of blocks changed per read = (block changes ) ( logical reads)

Recursive Call % 递归调用的比率;Recursive Call% = (recursive calls)/(user calls)

Rollback per transaction % 事务回滚比率。Rollback per transaction %= (rollback)/(transactions)
Rows per Sort平均每次排序涉及到的行数;Rows per Sort=(sorts(rows) ) ( sorts(disk) + sorts(memory))

注意这些Load Profile负载指标在本环节提供了2个维度per second和per transaction。

per Second: 主要是把快照内的时间值除以快照时间的秒数,例如在A快照中V$SYSSTAT视图反应 table scans (long tables) 这个指标是 100 ，
在B快照中V$SYSSTAT视图反应 table scans (long tables) 这个指标是 3700, 而A快照和B快照之间间隔了一个小时3600秒，则对于table scans (long tables) per second 就是 ( 3700- 100) 3600=1。

per transaction : 基于事务的维度，与per second相比是把除数从时间的秒数改为了该段时间内的事务数。这个维度的很大用户是用来识别应用特性的变化

若2个AWR性能报告中该维度指标出现了大幅变化，例如 redo size从本来per transaction 1k变化为 10k per transaction，则说明SQL业务逻辑肯定发生了某些变化。

注意AWR中的这些指标并不仅仅用来孤立地了解 Oracle数据库负载情况，实施调优工作。对于故障诊断例如HANG、Crash等，完全可以通过对比问题时段的性能报告和常规时间来对比，通过各项指标的对比往往可以找出病灶所在。

1-3 Instance Efficiency Percentages (Target 100%

Instance Efficiency Percentages (Target 100%)

Buffer Nowait %:	100.00	Redo NoWait %:	100.00
Buffer Hit %:	100.00	In-memory Sort %:	100.00
Library Hit %:	90.16	Soft Parse %:	96.72
Execute to Parse %:	0.93	Latch Hit %:	99.94
Parse CPU to Parse Elapsd %:	75.79	% Non-Parse CPU:	87.28

上述所有指标的目标均为100%，越大越好

80%以上 %Non-Parse CPU
90%以上 Buffer Hit%, In-memory Sort%, Soft Parse%
95%以上 Library Hit%, Redo Nowait%, Buffer Nowait%
98%以上 Latch Hit%

1、Buffer Nowait % session申请一个buffer(兼容模式)不等待的次数比例。需要访问buffer时立即可以访问的比率，

2、buffer HIT%: 高速缓存命中率，反应物理读和缓存命中间的纠结，但这个指标即便99% 也不能说明物理读等待少了

不合理的db_cache_size,或者是SGA自动管理ASMM Memory自动管理AMM下都可能因为db_cache_size过小引起大量的db file sequential scattered read等待事件；
此外与 buffer HIT%相关的指标值得关注的还有 table scans(long tables) 大表扫描这个统计项目、此外相关的栏目还有Buffer Pool Statistics 、Buffer Pool Advisory等

3、Redo nowait%: session在生成redo entry时不用等待的比例，redo相关的资源争用例如redo space request争用可能造成生成redo时需求等待。此项数据来源于v$sysstat中的(redo log space requests/redo entries)。一般来说10g以后不太用关注log_buffer参数的大小，需要关注是否有十分频繁的 log switch ；过小的redo logfile size 如果配合较大的SGA和频繁的commit提交都可能造成该问题。考虑增到redo logfile 的尺寸 : 1-2G 每个，10-15组都是合适的。同时考虑优化redo logfile和datafile 的I/O。

4、In-memory Sort%:这个指标因为它不计算workarea中所有的操作类型，纯粹在内存中完成的排序比例。

5、Library Hit%: library cache命中率，申请一个library cache object例如一个SQL cursor时，其已经在library cache中的比例。合理值：>95% ，ns

6、Soft Parse: 软解析比例，经典指标，合理值>95%
Soft Parse %是AWR中另一个重要的解析指标，该指标反应了快照时间内软解析次数和总解析次数 (soft+hard 软解析次数+硬解析次数)的比值，若该指标很低，那么说明了可能存在剧烈的hard parse硬解析，大量的硬解析会消耗更多的CPU时间片并产生解析争用(此时可以考虑使用cursor_sharing=FORCE)；
理论上我们总是希望 Soft Parse % 接近于100%，但并不是说100%的软解析就是最理想的解析状态，通过设置 session_cached_cursors参数和反复重用游标我们可以让解析来的更轻量级，即通俗所说的利用会话缓存游标实现的软软解析(soft soft parse)。

7、Execute to Parse%指标反映了执行解析比其公式为 1-(parse/execute) , 目标为100% 及接近于只执行而不解析。在oracle中解析往往是执行的先提工作，但是通过游标共享可以解析一次执行多次，

执行解析可能分成多种场景：
hard coding => 代码硬解析一次,执行一次，理论上其执行解析比为 1:1 ，则理论上Execute to Parse =0 极差，且soft parse比例也为0%
绑定变量但是仍软解析=》软解析一次，执行一次，这种情况虽然比前一种好但是执行解析比(这里的parse，包含了软解析和硬解析)仍是1:1，理论上Execute to Parse =0 极差，但是soft parse比例可能很高
使用静态SQL、动态绑定、session_cached_cursor、open cursors等技术实现的解析一次，执行多次，执行解析比为N:1，则Execute to Parse= 1- (1/N) 执行次数越多Execute to Parse越接近100%，这种是我们在OLTP环境中希望看到的。
通俗地说 soft parse% 反映了软解析率，而软解析在oracle中仍是较昂贵的操作，我们希望的是解析1次执行N次，如果每次执行均需要软解析，那么虽然soft parse%=100% 但是parse time仍可能是消耗DB TIME的大头。
Execute to Parse反映了执行解析比，Execute to Parse和soft parse% 都很低那么说明确实没有绑定变量，而如果 soft parse%接近99%而Execute to Parse 不足90% 则说明没有执行解析比低,需要通过静态SQL、动态绑定、session_cached_cursor、open cursors等技术减少软解析。

8、Latch Hit%: willing-to-wait latch闩申请不要等待的比例。

9、Parse CPU To Parse Elapsd:该指标反映了快照内解析CPU时间和总的解析时间的比值(Parse CPU Time/ Parse Elapsed Time)；若该指标水平很低，那么说明在整个解析过程中实际在CPU上运算的时间是很短的，而主要的解析时间都耗费在各种其他非空闲的等待事件上了(如latch:shared pool,row cache lock之类等)

10、%Non-Parse CPU非解析cpu比例,公式为 (DB CPU – Parse CPU)/DB CPU，

1-5 Top 10 Foreground Events by Total Wait Time

Top 10 Foreground Events by Total Wait Time

Event	Waits	Total Wait Time (sec)	Wait Avg(ms)	% DB time	Wait Class
DB CPU		3823.1		98.4
log file sync	10,933	101.5	9	2.6	Commit
cursor: pin S	3,299	23.4	7	.6	Concurrency
SQL*Net message to client	7,835,076	13.3	0	.3	Network
latch: shared pool	773	9.9	13	.3	Concurrency
direct path read	29,439	4.2	0	.1	User I/O
library cache: mutex X	681	3.1	5	.1	Concurrency
cursor: pin S wait on X	73	2.6	36	.1	Concurrency
latch: row cache objects	544	2.3	4	.1	Concurrency
SQL*Net more data to client	122,900	2	0	.1	Network

丰富的等待事件以足够的细节来描绘系统运行的性能瓶颈(Mysql梦寐以求的)
Waits : 该等待事件发生的次数

Times : 该等待事件消耗的总计时间,单位为秒,对于DB CPU而言是前台进程所消耗CPU时间片的总和

Avg Wait(ms) :该等待事件平均等待的时间,实际就是Times/Waits,单位ms，

Wait Class: 等待类型：
Concurrency,SystemI/O,UserI/O,Administrative,Other,Configuration,Scheduler,Cluster,Application,Idle,Network,Commit

常见的等待事件

=========================>

1、db file scattered read：

Avg wait time应当小于20ms如果数据库执行全表扫描或者是全索引扫描会执行 Multi block I/O ，此时等待物理I/O 结束会出现此等待事件。一般从应用程序(SQL），I/O 方面入手调整; 注意和index fast full scans (full) 以及 table scans结合起来一起看。

2、db file sequential read:

该等待事件Avg wait time平均单次等待时间应当小于20ms

db file sequential read单块读等待是一种最为常见的物理IO等待事件，这里的sequential指的是将数据块读入到相连的内存空间中，而不是指所读取的数据块是连续的。

3、latch free:

其实是未获得latch 而进入latch sleep

4、enq:XX 队列锁等待:

视乎不同的队列锁有不同的情况：

Oracle队列锁: Enqueue HW

enq: TX – row lock/index contention等待事件

enq: TT – contention等待事件

enq: JI – contention等待事件

enq: US – contention等待事件

enq: TM – contention等待事件

enq: RO fast object reuse等待事件

enq: HW – contention等待事件

5、free buffer waits：

是由于无法找到可用的buffer cache 空闲区域，需要等待DBWR 写入完成引起

一般是由于低效的sql、过小的buffer cache、DBWR 工作负荷过量引起，

6、buffer busy wait/ read by other session：

以上2个等待事件可以归为一起处理

7、write complete waits :

此类等待事件是由于DBWR 将脏数据写入数据文件，其他进程如果需要修改 buffer cache会引起此等待事件，一般是 I/O 性能问题或者是DBWR 工作负荷过量引起

8、control file parallel write：

频繁的更新控制文件会造成大量此类等待事件，如日志频繁切换，检查点经常发生，nologging 引起频繁的数据文件更改，I/O 系统性能缓慢。

9、log file sync：

此类等待时间是由于 LGWR 进程讲redo log buffer 写入redo log 中发生。如果此类事件频繁发生，可以判断为：commit 次数是否过多、I/O 系统问题、重做日志是否不必要被创建、redo log buffer是否过大

2-1 Time Model Statistics

Time Model Statistics

Total time in database user-calls (DB Time): 3886.1s
Statistics including the word "background" measure background process time, and so do not contribute to the DB time statistic
Ordered by % or DB time desc, Statistic name

Statistic Name	Time (s)	% of DB Time
DB CPU	3,823.09	98.38
sql execute elapsed time	1,382.61	35.58
parse time elapsed	698.98	17.99
hard parse elapsed time	340.29	8.76
PL/SQL execution elapsed time	8.17	0.21
connection management call elapsed time	1.19	0.03
hard parse (sharing criteria) elapsed time	0.56	0.01
PL/SQL compilation elapsed time	0.54	0.01
failed parse elapsed time	0.38	0.01
sequence load elapsed time	0.17	0.00
repeated bind elapsed time	0.02	0.00
DB time	3,886.11
background elapsed time	248.93
background cpu time	48.61

Time Model Statistics几个特别有用的时间指标：

parse time elapsed、hard parse elapsed time 结合起来看解析是否是主要矛盾，若是则重点是软解析还是硬解析

sequence load elapsed time sequence序列争用是否是问题焦点

PL/SQL compilation elapsed time PL/SQL对象编译的耗时

注意PL/SQL execution elapsed time 纯耗费在PL/SQL解释器上的时间。不包括花在执行和解析其包含SQL上的时间

connection management call elapsed time 建立数据库session连接和断开的耗时

failed parse elapsed time 解析失败，例如由于ORA-4031

hard parse(sharing criteria)elapsed time由于无法共享游标造成的硬解析

hard parse(bind mismatch)elapsed time由于bind type or bind size 不一致造成的硬解析
注意该时间模型中的指标存在包含关系所以Time Model Statistics加起来超过100%再正常不过

1) background elapsed time
2) background cpu time
3) RMAN cpu time (backup/restore)
1) DB time
2) DB CPU
2) connection management call elapsed time
2) sequence load elapsed time
2) sql execute elapsed time
2) parse time elapsed
3) hard parse elapsed time
4) hard parse (sharing criteria) elapsed time
5) hard parse (bind mismatch) elapsed time
3) failed parse elapsed time
4) failed parse (out of shared memory) elapsed time
2) PL/SQL execution elapsed time
2) inbound PL/SQL rpc elapsed time
2) PL/SQL compilation elapsed time
2) Java execution elapsed time
2) repeated bind elapsed time

2-2 Foreground Wait Class

Foreground Wait Class

s - second, ms - millisecond - 1000th of a second
ordered by wait time desc, waits desc
%Timeouts: value of 0 indicates value was < .5%. Value of null is truly 0
Captured Time accounts for 102.7% of Total DB time 3,886.11 (s)
Total FG Wait Time: 166.02 (s) DB CPU time: 3,823.09 (s)

Wait Class	Waits	%Time -outs	Total Wait Time (s)	Avg wait (ms)	%DB time
DB CPU			3,823		98.38
Commit	10,933	0	102	9	2.61
Concurrency	5,479	0	41	8	1.07
Network	7,965,502	0	15	0	0.40
User I/O	31,504	0	6	0	0.14
Application	339	0	2	5	0.04
Other	577	1	1	1	0.02
Configuration	2	100	0	6	0.00
System I/O	12	0	0	0	0.00

常见的WAIT_CLASS类型
---------------------------------------
Concurrency
User I/O
System I/O
Administrative
Other
Configuration
Scheduler
Cluster
Application
Queueing
Idle
Network
Commit

Wait Class: 等待事件的类型。

Waits: 该类型所属等待事件在快照时间内的等待次数

%Time Out 等待超时的比率，未超时次数/waits * 100 (%)

Total Wait Time: 该类型所属等待事件总的耗时，单位为秒

Avg Wait(ms) : 该类型所属等待事件的平均单次等待时间，单位为ms ，实际这个指标对commit 和 user i/o 以及system i/o类型有点意义

Other 类型，遇到该类型等待事件的话常见的原因是Oracle Bug或者网络、I/O存在问题，建议提交MOS

Concurrency 类型并行争用类型的等待事件,典型的如 latch: shared pool、latch: library cache、row cache lock、library cache pin/lock

Cluster 类型为Real Application Cluster RAC环境中的等待事件，需要注意的是如果启用了RAC option，那么即使你的集群中只启动了一个实例，那么该实例也可能遇到 Cluster类型的等待事件, 例如gc buffer busy

System I/O 主要是后台进程维护数据库所产生的I/O，例如control file parallel write 、log file parallel write、db file parallel write。

User I/O 主要是前台进程做了一些I/O操作，并不是说后台进程不会有这些等待事件。典型的如db file sequential/scattered read、direct path read

Configuration 由于配置引起的等待事件，例如日志切换的log file switch

Application 应用造成的等待事件，例如enq: TM – contention和enq: TX – row lock contention；Oracle认为这是由于应用设计不当造成的等待事件，但实际这些Application class 等待可能受到 Concurrency、Cluster、System I/O 、User I/O等多种类型等待的影响，例如本来commit只要1ms ，则某一行数据仅被锁定1ms，但由于commit变慢从而释放行锁变慢，引发大量的enq: TX – row lock contention等待事件。

Network : 网络类型的等待事件例如 SQL*Net more data to client 、SQL*Net more data to dblink

Idle 空闲等待事件，最为常见的是rdbms ipc message

SQL*Net message from client 等待SQL*NET传来信息

2-3 Foreground Wait Events

Foreground Wait Events

s - second, ms - millisecond - 1000th of a second
Only events with Total Wait Time (s) >= .001 are shown
ordered by wait time desc, waits desc (idle events last)
%Timeouts: value of 0 indicates value was < .5%. Value of null is truly 0

Event	Waits	%Time -outs	Total Wait Time (s)	Avg wait (ms)	Waits txn	% DB time
log file sync	10,933	0	102	9	0.01	2.61
cursor: pin S	3,299	0	23	7	0.00	0.60
SQL*Net message to client	7,835,076	0	13	0	8.50	0.34
latch: shared pool	773	0	10	13	0.00	0.26
direct path read	29,439	0	4	0	0.03	0.11
library cache: mutex X	681	0	3	5	0.00	0.08
cursor: pin S wait on X	73	0	3	36	0.00	0.07
latch: row cache objects	544	0	2	4	0.00	0.06
SQL*Net more data to client	122,900	0	2	0	0.13	0.05
enq: KO - fast object checkpoint	23	0	1	48	0.00	0.03
db file sequential read	1,608	0	1	1	0.00	0.03
latch free	545	0	0	1	0.00	0.01
direct path sync	15	0	0	18	0.00	0.01
kksfbc child completion	4	100	0	51	0.00	0.01
enq: TX - row lock contention	8	0	0	25	0.00	0.01
SQL*Net break/reset to client	306	0	0	1	0.00	0.01
SQL*Net more data from client	7,526	0	0	0	0.01	0.00
local write wait	7	0	0	4	0.00	0.00
Disk file operations I/O	177	0	0	0	0.00	0.00
enq: RO - fast object reuse	2	0	0	11	0.00	0.00
latch: cache buffers chains	53	0	0	0	0.00	0.00
reliable message	16	0	0	1	0.00	0.00
undo segment extension	2	100	0	6	0.00	0.00
cursor: mutex S	1	0	0	11	0.00	0.00
buffer busy waits	24	0	0	0	0.00	0.00
read by other session	2	0	0	3	0.00	0.00
direct path write	243	0	0	0	0.00	0.00
db file parallel read	1	0	0	3	0.00	0.00
latch: session allocation	10	0	0	0	0.00	0.00
latch: In memory undo latch	12	0	0	0	0.00	0.00
library cache lock	2	0	0	1	0.00	0.00
library cache load lock	6	0	0	0	0.00	0.00
latch: enqueue hash chains	1	0	0	1	0.00	0.00
SQL*Net message from client	7,835,081	0	261,357	33	8.50
jobq slave wait	7,496	100	3,763	502	0.01

2-4 Background Wait Events

Background Wait Events

ordered by wait time desc, waits desc (idle events last)
Only events with Total Wait Time (s) >= .001 are shown
%Timeouts: value of 0 indicates value was < .5%. Value of null is truly 0

Event	Waits	%Time -outs	Total Wait Time (s)	Avg wait (ms)	Waits txn	% bg time
log file parallel write	11,570	0	104	9	0.01	41.67
db file async I/O submit	2,497	0	75	30	0.00	30.27
control file parallel write	1,576	0	17	11	0.00	6.77
os thread startup	139	0	8	60	0.00	3.35
db file sequential read	175	0	0	2	0.00	0.13
latch: shared pool	17	0	0	16	0.00	0.11
control file sequential read	4,168	0	0	0	0.00	0.06
direct path sync	1	0	0	104	0.00	0.04
log file sync	2	0	0	28	0.00	0.02
ADR block file read	16	0	0	2	0.00	0.01
ADR block file write	5	0	0	3	0.00	0.01
reliable message	23	0	0	0	0.00	0.00
asynch descriptor resize	245	100	0	0	0.00	0.00
Disk file operations I/O	66	0	0	0	0.00	0.00
LGWR wait for redo copy	32	0	0	0	0.00	0.00
direct path write	23	0	0	0	0.00	0.00
rdbms ipc message	30,297	62	61,071	2016	0.03
DIAG idle wait	7,192	100	7,214	1003	0.01
smon timer	15	73	3,755	250334	0.00
shared server idle wait	121	100	3,631	30010	0.00
Streams AQ: qmn coordinator idle wait	266	48	3,613	13583	0.00
Streams AQ: qmn slave idle wait	129	0	3,613	28008	0.00
pmon timer	1,201	100	3,609	3005	0.00
Space Manager: slave idle wait	722	100	3,608	4997	0.00
dispatcher timer	60	100	3,601	60012	0.00
SQL*Net message from client	532	0	0	1	0.00
class slave wait	19	0	0	0	0.00

2-5 Operating System Statistics

Operating System Statistics

*TIME statistic values are diffed. All others display actual values. End Value is displayed if different
ordered by statistic type (CPU Use, Virtual Memory, Hardware Config), Name

Statistic	Value	End Value
BUSY_TIME	389,214
IDLE_TIME	5,201,491
IOWAIT_TIME	11,255
NICE_TIME	0
SYS_TIME	155,118
USER_TIME	227,910
LOAD	0	1
RSRC_MGR_CPU_WAIT_TIME	0
VM_IN_BYTES	0
VM_OUT_BYTES	0
PHYSICAL_MEMORY_BYTES	33,805,135,872
NUM_CPUS	16
NUM_CPU_CORES	16
NUM_CPU_SOCKETS	4
GLOBAL_RECEIVE_SIZE_MAX	4,194,304
GLOBAL_SEND_SIZE_MAX	1,048,576
TCP_RECEIVE_SIZE_DEFAULT	87,380
TCP_RECEIVE_SIZE_MAX	4,194,304
TCP_RECEIVE_SIZE_MIN	4,096
TCP_SEND_SIZE_DEFAULT	16,384
TCP_SEND_SIZE_MAX	4,194,304
TCP_SEND_SIZE_MIN	4,096

Operating System Statistics 操作系统统计信息, TIME相关的指标单位均为百分之一秒

NUM_CPU_SOCKETS:物理CPU的数目

NUM_CPU_CORES:CPU的核数
NUM_CPUS:逻辑CPU的数目
SYS_TIME:在内核态被消耗掉的CPU时间片，单位为百分之一秒
USER_TIME:在用户态被消耗掉的CPU时间片，单位为百分之一秒
BUSY_TIME:Busy_Time=SYS_TIME+USER_TIME 消耗的CPU时间片，单位为百分之一秒
AVG_BUSY_TIME:AVG_BUSY_TIME= BUSY_TIME/NUM_CPUS
IDLE_TIME 空闲的CPU时间片，单位为百分之一秒
OS_CPU_WAIT_TIME:进程等OS调度的时间
IOWAIT_TIME:所有CPU花费在等待I/O完成上的时间单位为百分之一秒

2-6 Service Statistcs

Service Statistics

ordered by DB Time

Service Name	DB Time (s)	DB CPU (s)	Physical Reads (K)	Logical Reads (K)
SYS$USERS	3,886	3,823	1,428	81,868
SYS$BACKGROUND	0	0	0	63
cufsoa	0	0	0	0
cufsoaXDB	0	0	0	0

Service Name 对应的服务名 (v$services)，

SYS$BACKGROUND代表后台进程，

SYS$USERS一般是系统用户登录

DB TIME (s): 本服务名所消耗的DB TIME时间，单位为秒

DB CPU(s): 本服务名所消耗的DB CPU 时间，单位为秒

Physical Reads : 本服务名所消耗的物理读

Logical Reads : 本服务所消耗的逻辑读

2-7 Service Wait Class Stats

Service Wait Class Stats

Wait Class info for services in the Service Statistics section.
Total Waits and Time Waited displayed for the following wait classes: User I/O, Concurrency, Administrative, Network
Time Waited (Wt Time) in seconds

Service Name	User I/O Total Wts	User I/O Wt Time	Concurcy Total Wts	Concurcy Wt Time	Network Total Wts	Network Wt Time
SYS$USERS	31504	6	5479	41	7965512	15
SYS$BACKGROUND	264	0	148	8	0	0
cufsoa	0	0	0	0	0	0

User I/O Total Wts : 对应该服务名下用户I/O类等待的总的次数

User I/O Wt Time :对应该服务名下用户I/O累等待的总时间,单位为 1/100秒

Concurcy Total Wts: 对应该服务名下 Concurrency 类型等待的总次数

Concurcy Wt Time :对应该服务名下 Concurrency 类型等待的总时间，单位为 1/100秒

Admin Total Wts: 对应该服务名下Admin 类等待的总次数

Admin Wt Time: 对应该服务名下Admin类等待的总时间，单位为1/100秒

Network Total Wts : 对应服务名下Network类等待的总次数

Network Wt Time：对应服务名下Network类等待的总事件，单位为1/100秒

2-8 Host CPU

Host CPU

CPUs	Cores	Sockets	Load Average Begin	Load Average End	%User	%System	%WIO	%Idle
16	16	4	0.48	0.62	4.1	2.8	0.2	93.0

Load Average begin/end值代表每个CPU的大致运行队列大小。

2-8 Instance CPU

Instance CPU

%Total CPU	%Busy CPU	%DB time waiting for CPU (Resource Manager)
6.9	99.5	0.0

%Busy CPU，该实例所使用的Cpu占总的被使用CPU的比例 % of busy CPU for Instance

例如共4个逻辑CPU，其中3个被完全使用，3个中的1个完全被该实例使用，则%Total CPU= ¼ =25%，而%Busy CPU= 1/3= 33%
当CPU高时一般看%Busy CPU可以确定CPU到底是否是本实例消耗的，还是主机上其他程序

3 TOP SQL

3-1 SQL ordered by Elapsed Time ，

SQL ordered by Elapsed Time

Resources reported for PL/SQL code includes the resources used by all SQL statements called by the code.
% Total DB Time is the Elapsed Time of the SQL statement divided into the Total Database Time multiplied by 100
%Total - Elapsed Time as a percentage of Total DB time
%CPU - CPU Time as a percentage of Elapsed Time
%IO - User I/O Time as a percentage of Elapsed Time
Captured SQL account for 13.9% of Total DB Time (s): 3,886
Captured PL/SQL account for 0.1% of Total DB Time (s): 3,886

Elapsed Time (s)	Executions	Elapsed Time per Exec (s)	%Total	%CPU	%IO	SQL Id	SQL Module	SQL Text
192.02	916,682	0.00	4.94	46.61	0.00	7gwxpwru0czqw
27.05	59	0.46	0.70	98.60	0.00	4nwgd6fbmbkm0

注意对于PL/SQL，SQL Statistics不仅会体现该PL/SQL的执行情况，还会包括该PL/SQL包含的SQL语句的情况。

对于Top SQL很有必要一探究竟

Elapsed Time (s): 该SQL累计运行所消耗的时间，

Executions : 该SQL在快照时间内总计运行的次数；注意对于在快照时间内还没有执行完的SQL不计为一次，所以如果看到executions=0而又是TOP SQL,则很有可能是因为该SQL运行较旧还没执行完，需要特别关注一下。

Elapsed Time per Exec (s)：平均每次执行该SQL耗费的时间，对于OLTP类型的SELECT/INSERT/UPDATE/DELETE而言平均单次执行时间应当非常短，
如0.1秒或者更短才能满足其业务需求，如果这类轻微的OLTP操作单次也要几秒钟的话，是无法满足对外业务的需求的；如果这些操作也变得很慢，则会出现大量事务阻塞，系统负载升高，DB TIME急剧上升的现象。对于OLTP数据库而言如果执行计划稳定，那么这些OLTP操作的性能应当是稳定的，但是一旦某个因素发生变化,例如存储的明显变慢、内存换页的大量出现时则上述的这些transaction操作很可能成数倍到几十倍的变慢，这将让此事务系统短期内不可用。

%Total 该SQL所消耗的时间占总的DB Time的百分比，即 (SQL Elapsed Time Total DB TIME)

% CPU 该SQL 所消耗的CPU时间占该SQL消耗的时间里的比例，

%IO 该SQL 所消耗的I/O时间占该SQL消耗的时间里的比例，该指标说明了该语句是否是I/O敏感的

SQL Id : 通过计算SQL 文本获得的SQL_ID ，不同的SQL文本必然有不同的SQL_ID，对于10g~11g而言只要SQL文本不变那么在数据库之间该SQL 对应的SQL_ID应当不不变的， 12c中修改了SQL_ID的计算方法

3-2 SQL ordered by CPU Time

SQL ordered by CPU Time

Resources reported for PL/SQL code includes the resources used by all SQL statements called by the code.
%Total - CPU Time as a percentage of Total DB CPU
%CPU - CPU Time as a percentage of Elapsed Time
%IO - User I/O Time as a percentage of Elapsed Time
Captured SQL account for 10.1% of Total CPU Time (s): 3,823
Captured PL/SQL account for 0.0% of Total CPU Time (s): 3,823

CPU Time (s)	Executions	CPU per Exec (s)	%Total	Elapsed Time (s)	%CPU	%IO	SQL Id	SQL Module	SQL Text
89.51	916,682	0.00	2.34	192.02	46.61	0.00	7gwxpwru0czqw
26.67	59	0.45	0.70	27.05	98.60	0.00	4nwgd6fbmbkm0

CPU TIME : 该SQL在快照时间内累计执行所消耗的CPU时间片，单位为s

Executions : 该SQL在快照时间内累计执行的次数

CPU per Exec (s) ：该SQL 平均单次执行所消耗的CPU时间，

%Total : 该SQL累计消耗的CPU时间占该时段总的DB CPU的比例，

% CPU 该SQL 所消耗的CPU时间占该SQL消耗的时间里的比例，该指标说明了该语句是否是CPU敏感的

%IO 该SQL 所消耗的I/O时间占该SQL消耗的时间里的比例，该指标说明了该语句是否是I/O敏感的

3-3 Buffer Gets SQL ordered by Gets

SQL ordered by Reads

%Total - Physical Reads as a percentage of Total Disk Reads
%CPU - CPU Time as a percentage of Elapsed Time
%IO - User I/O Time as a percentage of Elapsed Time
Total Disk Reads: 1,427,765
Captured SQL account for 0.0% of Total

Physical Reads	Executions	Reads per Exec	%Total	Elapsed Time (s)	%CPU	%IO	SQL Id	SQL Module	SQL Text
0	3,218	0.00	0.00	0.78	37.01	0.00	074w4jr8a9fu8
0	33,180	0.00	0.00	7.55	63.61	0.00	09tx5h4ngu7va

注意 buffer gets 逻辑读是消耗CPU TIME的重要源泉，但并不是说消耗CPU TIME的只有buffer gets。

大多数情况下 SQL order by CPU TIME和SQL order by buffers gets 2个部分的TOP SQL 及其排列顺序都是一样的，此种情况说明消耗最多buffer gets的就是消耗最多CPU 的SQL ，如果我们希望降低系统的CPU使用率，那么只需要调优SQL 降低buffer gets 即可。

但也并不是100%的情况都是如此，CPU TIME的消耗者还包括函数运算、PL/SQL 控制、Latch Mutex 的Spin等等，所以SQL order by CPU TIME 和 SQL order by buffers gets 2个部分的TOP SQL 完全不一样也是有可能的，

Buffer Gets : 该SQL在快照时间内累计运行所消耗的buffer gets，包括了consistent read 和 current read

Executions : 该SQL在快照时间内累计执行的次数

Gets per Exec : 该SQL平均单次的buffer gets ，对于事务型transaction操作而言一般该单次buffer gets小于2000

% Total 该SQL累计运行所消耗的buffer gets占总的db buffer gets的比率，

3-4 Physical Reads SQL ordered by Reads

SQL ordered by Physical Reads (UnOptimized)

UnOptimized Read Reqs = Physical Read Reqts - Optimized Read Reqs
%Opt - Optimized Reads as percentage of SQL Read Requests
%Total - UnOptimized Read Reqs as a percentage of Total UnOptimized Read Reqs
Total Physical Read Requests: 31,234
Captured SQL account for 0.0% of Total
Total UnOptimized Read Requests: 31,234
Captured SQL account for 0.0% of Total
Total Optimized Read Requests: 1
Captured SQL account for 0.0% of Total

UnOptimized Read Reqs	Physical Read Reqs	Executions	UnOptimized Reqs per Exec	%Opt	%Total	SQL Id	SQL Module	SQL Text
12	12	1	12.00	0.00	0.04	gjm43un5cy843
0	0	3,218	0.00		0.00	074w4jr8a9fu8

Physical reads : 该SQL累计运行所消耗的物理读

Executions : 该SQL在快照时间内累计执行的次数

Reads per Exec : 该SQL单次运行所消耗的物理读，对于OLTP transaction 类型的操作而言单次一般不超过100

%Total : 该SQL 累计消耗的物理读占该时段总的物理读的比例，

3-5 Executions SQL ordered by Executions

SQL ordered by Executions

%CPU - CPU Time as a percentage of Elapsed Time
%IO - User I/O Time as a percentage of Elapsed Time
Total Executions: 2,315,874
Captured SQL account for 52.1% of Total

Executions	Rows Processed	Rows per Exec	Elapsed Time (s)	%CPU	%IO	SQL Id	SQL Module	SQL Text
916,682	916,316	1.00	192.02	46.6	0	7gwxpwru0czqw
51,293	0	0.00	15.91	40.2	0	fdywauggm8byz

按照执行次数来排序的话，也是性能报告对比时一个重要的参考因素，因为如果TOP SQL的执行次数有明显的增长，那么性能问题的出现也是意料之中的事情了。当然执行次数最多的，未必便是对性能影响最大的TOP SQL

Executions : 该SQL在快照时间内累计执行的次数

Rows Processed：该SQL在快照时间内累计执行所处理的总行数
Rows per Exec：SQL平均单次执行所处理的行数，这个指标在诊断一些数据问题造成的SQL性能问题时很有用

3-6 Parse Calls SQL ordered by Parse Calls

SQL ordered by Parse Calls

Total Parse Calls: 2,294,317
Captured SQL account for 52.6% of Total

Parse Calls	Executions	% Total Parses	SQL Id	SQL Module	SQL Text
916,829	916,682	39.96	7gwxpwru0czqw		select companyname from licens...
51,293	51,293	2.24	fdywauggm8byz		select * from WorkFlowPlanSet ...

Parse Calls : 解析调用次数，与上文的 Load Profile中的Parse数一样包括软解析soft parse和硬解析hard parse

Executions : 该SQL在快照时间内累计执行的次数

%Total Parses : 本SQL 解析调用次数占该时段数据库总解析次数的比率

3-7 SQL ordered by Sharable Memory

Shareable Mem(b): SQL对象所占用的共享内存使用量

Executions : 该SQL在快照时间内累计执行的次数

%Total : 该SQL 对象锁占共享内存占总的共享内存的比率

3-8 SQL ordered by Version Count

SQL ordered by Version Count

Only Statements with Version Count greater than 20 are displayed

Version Count	Executions	SQL Id	SQL Module	SQL Text
22	1	6wm3n4d7bnddg		SELECT source, (case when tim...

Executions : 该SQL在快照时间内累计执行的次数

Hash Value : 共享SQL 的哈希值

4 Instance Activity Stats

4-1 Other by statistic name

Other Instance Activity Stats

Ordered by statistic name

Statistic	Total	per Second	per Trans
Batched IO (bound) vector count	0	0.00	0.00
Batched IO (full) vector count	0	0.00	0.00
Batched IO (space) vector count	0	0.00	0.00
Batched IO block miss count	3	0.00	0.00
Batched IO buffer defrag count	0	0.00	0.00
Batched IO double miss count	1	0.00	0.00
Batched IO same unit count	0	0.00	0.00
Batched IO single block count	0	0.00	0.00
Batched IO slow jump count	0	0.00	0.00
Batched IO vector block count	3	0.00	0.00
Batched IO vector read count	1	0.00	0.00
Block Cleanout Optim referenced	6	0.00	0.00
CCursor + sql area evicted	321	0.09	0.00
CPU used by this session	189,051	52.39	0.21
CPU used when call started	187,601	51.99	0.20
CR blocks created	261	0.07	0.00
Cached Commit SCN referenced	567,554	157.29	0.62
Commit SCN cached	74	0.02	0.00
DBWR checkpoint buffers written	16,185	4.49	0.02
DBWR checkpoints	26	0.01	0.00
DBWR object drop buffers written	6	0.00	0.00
DBWR revisited being-written buffer	0	0.00	0.00
DBWR tablespace checkpoint buffers written	157	0.04	0.00
DBWR thread checkpoint buffers written	0	0.00	0.00
DBWR transaction table writes	122	0.03	0.00
DBWR undo block writes	6,538	1.81	0.01
HSC Heap Segment Block Changes	143,725	39.83	0.16
HSC IDL Compressed Blocks	0	0.00	0.00
Heap Segment Array Inserts	2,807	0.78	0.00
Heap Segment Array Updates	4,642	1.29	0.01
IMU CR rollbacks	218	0.06	0.00
IMU Flushes

这里每一个指标都代表一种数据库行为的活跃度，

虽然这些指标均只有Total、per Second每秒、 per Trans每事务三个维度，但对诊断问题十分有用。

例如：

1、例如当 Top Event 中存在direct path read为Top 等待事件，则需要分清楚是对普通堆表的direct read还是由于大量LOB读造成的direct path read，这个问题可以借助 table scans (direct read)、table scans (long tables)、physical reads direct 、physical reads direct (lob) 、
physical reads direct temporary几个指标来分析，假设 physical reads direct >> 远大于 physical reads direct (lob)+physical reads direct temporary ，且有较大的table scans (direct read)、table scans (long tables) (注意这2个指标代表的是扫描表的次数不同于上面的phsical reads 的单位为块数*次数)，则说明了是大表扫描引起的direct path read。

2、例如当 Top Event中存在enq Tx:index contention等待事件，则需要分析root node splits 、branch node splits 、leaf node 90-10 splits 、leaf node splits 、failed probes on index block rec 几个指标

4-2 Statistics with absolute values

Instance Activity Stats - Absolute Values

Statistics with absolute values (should not be diffed)

Statistic	Begin Value	End Value
logons current	85	80
opened cursors current	793	275
session cursor cache count	17,245,584	17,263,174
session pga memory	373,767,091,272	373,863,965,728
session pga memory max	429,564,886,056	429,369,395,424
session uga memory	283,783,601,784	283,855,681,528
session uga memory max	758,640,598,976	759,257,627,368

Instance Activity Stats – Absolute Values是显示快照起点和终点的一些指标的绝对值

logon current ：当前时间点的登录数

opened cursors current ：当前打开的游标数

session cursor cache count ：当前存在的session缓存游标数

5 IO 统计

5-1 Tablespace IO Stats 基于表空间分组的IO信息

Tablespace IO Stats

ordered by IOs (Reads + Writes) desc

Tablespace	Reads	Av Rds/s	Av Rd(ms)	Av Blks/Rd	1-bk Rds/s	Av 1-bk Rd(ms)	Writes	Writes avg/s	Buffer Waits	Av Buf Wt(ms)
USERS	24,117	7	0.02	26.68	4,267	5.18	0	1	4	0.00
1	6,391	2	0.00	122.52	714	0.06	0	0	19	1.05
1	271	0	1.11	1.07	1,673	0.07	1	0	2	0.00
SYSAUX	138	0	1.74	1.00	904	0.04	2	0	0	0.00
UNDOTBS1	0	0	0.00	0.00	652	0.00	0	0	4	2.50
SYSTEM	299	0	0.94	1.00	122	0.08	1	0	1	0.00
TEMP	6	0	0.00	1.00	1	0.00	0	0	0	0.00

reads : 指该表空间上发生的物理读的次数(单位不是块，而是次数)

Av Reads/s : 指该表空间上平均每秒的物理读次数 (单位不是块，而是次数)

Av Rd(ms): 指该表空间上每次读的平均读取延迟

Av Blks/Rd: 指该表空间上平均每次读取的块数目，因为一次物理读可以读多个数据块；如果Av Blks/Rd>>1则可能系统有较多db file scattered read 可能是诊断FULL TABLE SCAN或FAST FULL INDEX SCAN，需要关注table scans (long tables) 和index fast full scans (full) 2个指标

Writes : 该表空间上发生的物理写的次数 ; 对于那些Writes总是等于0的表空间不妨了解下是否数据为只读，如果是可以通过read only tablespace来解决 RAC中的一些性能问题。

Av Writes/s : 指该表空间上平均每秒的物理写次数

buffer Waits: 该表空间上发生buffer busy waits和read by other session的次数( 9i中buffer busy waits包含了read by other session)。

Av Buf Wt(ms): 该表空间上发生buffer Waits的平均等待时间，单位为ms

5-2 File I/O

File IO Stats

ordered by Tablespace, File

Tablespace	Filename	Reads	Av Rds/s	Av Rd(ms)	Av Blks/Rd	1-bk Rds/s	Av 1-bk Rd(ms)	Writes	Writes avg/s	Buffer Waits	Av Buf Wt(ms)
1	01.dbf	6,391	2	0.00	122.52	0	0.09	714	0	19	1.05
1	01.dbf	271	0	1.11	1.07	0	1.09	1,673	0	2	0.00
SYSAUX	01.dbf	138	0	1.74	1.00	0	1.74	904	0	0	0.00
SYSTEM	01.dbf	299	0	0.94	1.00	0	0.97	122	0	1	0.00
TEMP	01.dbf	6	0	0.00	1.00	0	0.00	1	0	0
UNDOTBS1	01.dbf	0	0			0		652	0	4	2.50
USERS	01.dbf	24,117	7	0.02	26.68	5	0.02	4,267	1	4	0.00

Tablespace 表空间名

FileName 数据文件的路径

Reads: 该数据文件上累计发生过的物理读次数，不是块数

Av Reads/s: 该数据文件上平均每秒发生过的物理读次数，不是块数

Av Rd(ms): 该数据文件上平均每次物理读取的延迟，单位为ms

Av Blks/Rd: 该数据文件上平均每次读取涉及到的块数，OLTP环境该值接近 1

Writes : 该数据文件上累计发生过的物理写次数，不是块数

Av Writes/s: 该数据文件上平均每秒发生过的物理写次数，不是块数

buffer Waits: 该数据文件上发生buffer busy waits和read by other session的次数

Av Buf Wt(ms): 该数据文件上发生buffer Waits的平均等待时间，单位为ms

若某个表空间上有较高的IO负载，则有必要分析一下是否其所属的数据文件上的IO 较为均匀还是存在倾斜，是否需要结合存储特征来将数据均衡分布到不同磁盘上的数据文件上，以优化 I/O

6 缓冲池统计 Buffer Pool Statistics

Buffer Pool Statistics

Standard block size Pools D: default, K: keep, R: recycle
Default Pools for other block sizes: 2k, 4k, 8k, 16k, 32k

P	Number of Buffers	Pool Hit%	Buffer Gets	Physical Reads	Physical Writes	Free Buff Wait	Writ Comp Wait	Buffer Busy Waits
D	354,600	100	80,349,220	1,603	16,200	0	0	30

P: pool池的名字

D: 默认的缓冲池 default buffer pool ,

K : Keep Pool ,

R: Recycle Pool ;

2k 4k 8k 16k 32k: 代表各种非标准块大小的缓冲池

Number of buffers: 实际的缓冲块数目，约等于池的大小/池的块大小

Pool Hit % : 该缓冲池的命中率

Buffer Gets: 对该缓冲池的中块的访问次数包括consistent gets 和db block gets

Physical Reads: 该缓冲池Buffer Cache引起了多少物理读，其实是physical reads cache ，单位为块数*次数

Physical Writes ：该缓冲池中Buffer cache被写的物理写，其实是physical writes from cache，单位为块数*次数

Free Buffer Waits: 等待空闲缓冲的次数，可以看做该buffer pool 发生free buffer waits 等待的次数

Write Comp Wait: 等待DBWR写入脏buffer到磁盘的次数，可以看做该buffer pool发生write complete waits等待的次数

Buffer Busy Waits: 该缓冲池发生buffer busy wait 等待的次数

7-1 Checkpoint Activity 检查点

Checkpoint Activity

Total Physical Writes: 16,478

MTTR Writes	Log Size Writes	Log Ckpt Writes	Other Settings Writes	Autotune Ckpt Writes	Thread Ckpt Writes
0	0	0	0	16,028	0

Instance Recovery Stats 实例恢复

Instance Recovery Stats

B: Begin Snapshot, E: End Snapshot

	Targt MTTR (s)	Estd MTTR (s)	Recovery Estd IOs	Actual RedoBlks	Target RedoBlks	Log Sz RedoBlks	Log Ckpt Timeout RedoBlks	Log Ckpt Interval RedoBlks	Opt Log Sz(M)	Estd RAC Avail Time
B	0	23	1056	4461	39600	3397383	39600
E	0	23	1505	16588	74323	3397383	74323

MTTR Writes : 为了满足FAST_START_MTTR_TARGET指定的MTTR值而做出的物理写 WRITES_MTTR

Log Size Writes ：由于最小的redo log file而做出的物理写 WRITES_LOGFILE_SIZE

Log Ckpt writes:由于 LOG_CHECKPOINT_INTERVAL 和 LOG_CHECKPOINT_TIMEOUT 驱动的增量检查点而做出的物理写 WRITES_LOG_CHECKPOINT_SETTINGS

Other Settings Writes ：由于其他设置(例如FAST_START_IO_TARGET）而引起的物理写， WRITES_OTHER_SETTINGS

Autotune Ckpt Writes : 由于自动调优检查点而引起的物理写， WRITES_AUTOTUNE

Thread Ckpt Writes ：由于thread checkpoint而引起的物理写，WRITES_FULL_THREAD_CKPT

B 代表开始点， E 代表结尾

Targt MTTR (s) : 目标MTTR (mean time to recover)意为有效恢复时间，单位为秒。

estimated_mttr (s): 当前基于脏buffer和重做日志块的数量，而评估出的有效恢复时间。它的估算告诉用户以当下系统的负载若发生实例crash，则需要多久时间来做crash recovery的前滚操作，之后才能打开数据库。

Recovery Estd IOs ：实际是当前buffer cache中的脏块数量，一旦实例崩溃这些脏块要被前滚

Actual RedoBlks ：当前实际需要恢复的redo重做块数量

Target RedoBlks :是Log Sz RedoBlks、Log Ckpt Timeout RedoBlks、 Log Ckpt Interval RedoBlks 三者的最小值

Log Sz RedoBlks : 代表必须在log file switch日志切换之前完成的 checkpoint 中涉及到的redo block，也叫max log lag；

Log Ckpt Timeout RedoBlks ：为了满足LOG_CHECKPOINT_TIMEOUT 所需要处理的redo block数，lag for checkpoint timeout ；

Log Ckpt Interval RedoBlks ：为了满足LOG_CHECKPOINT_INTERVAL 所需要处理的redo block数， lag for checkpoint interval；

Opt Log Sz(M) : 基于FAST_START_MTTR_TARGET 而估算出来的redo logfile 的大小，单位为MB 。

7-2 Buffer Pool Advisory 缓冲池建议

Buffer Pool Advisory

Only rows with estimated physical reads >0 are displayed
ordered by Block Size, Buffers For Estimate

P	Size for Est (M)	Size Factor	Buffers (thousands)	Est Phys Read Factor	Estimated Phys Reads (thousands)	Est Phys Read Time	Est %DBtime for Rds
D	288	0.10	35	96.40	2,217,689	1	466300.00
D	576	0.20	71	31.73	729,999	1	153419.00
D	864	0.30	106	13.10	301,470	1	63293.00
D	1,152	0.40	142	7.56	173,942	1	36472.00
D	1,440	0.49	177	4.62	106,305	1	22247.00
D	1,728	0.59	213	3.27	75,245	1	15715.00
D	2,016	0.69	248	2.58	59,336	1	12369.00
D	2,304	0.79	284	2.03	46,738	1	9720.00
D	2,592	0.89	319	1.46	33,610	1	6958.00
D	2,880	0.99	355	1.02	23,504	1	4833.00
D	2,912	1.00	359	1.00	23,006	1	4728.00
D	3,168	1.09	390	0.74	16,975	1	3460.00
D	3,456	1.19	426	0.62	14,344	1	2907.00
D	3,744	1.29	461	0.56	12,781	1	2578.00
D	4,032	1.38	496	0.50	11,420	1	2292.00
D	4,320	1.48	532	0.46	10,541	1	2107.00
D	4,608	1.58	567	0.43	9,966	1	1986.00
D	4,896	1.68	603	0.42	9,587	1	1906.00
D	5,184	1.78	638	0.41	9,326	1	1851.00
D	5,472	1.88	674	0.40	9,124	1	1809.00
D	5,760	1.98	709	0.38	8,778	1	1736.00

缓冲池的颗粒大小

P 指缓冲池的名字可能包括有

D default buffer pool ,

K Keep Pool ,

R recycle Pool

Size For Est(M): 指以该尺寸的buffer pool作为评估的对象，一般是目前current size的 10% ~ 200%，以便了解 buffer pool增大 ~减小对物理读的影响

Size Factor : 尺寸因子，只对应buffer pool大小对当前设置的比例因子

Buffers (thousands) ：指这个buffer pool 尺寸下的buffer数量，要乘以1000才是实际值

Est Phys Read Factor ：评估的物理读因子，

Estimated Phys Reads (thousands)：评估的物理读数目，要乘以1000才是实际值，显然不同尺寸的buffer pool对应不同的评估的物理读数目

Est Phys Read Time ：评估的物理读时间

Est %DBtime for Rds：评估的物理读占DB TIME的比率

我们看buffer pool advisory 一般有2个目的：

在物理读较多的情况下，希望通过增加buffer pool 大小来缓解物理读等待，这是我们关注Size Factor > 1的buffer pool尺寸是否能共有效减少Est Phys Read Factor，如果Est Phys Read Factor随着Size Factor增大而显著减少，那么说明增大buffer cache 是可以有效减少物理读的。在内存紧张的情况下，希望从buffer pool中匀出部分内存来移作他用，但是又不希望 buffer cache变小导致物理读增多性能下降，则此时观察Est Phys Read Factor 是否随着Size Factor 减小而显著增大，如果不是则说明减少部分buffer cache 不会导致物理读大幅增加，也就可以安心减少buffer cache

7-3 PGA Aggr Summary

PGA Aggr Summary

PGA cache hit % - percentage of W/A (WorkArea) data processed only in-memory

PGA Cache Hit %	W/A MB Processed	Extra W/A MB Read/Written
100.00	9,696	0

PGA Cache Hit % : 指W/A WorkArea工作区的数据仅在内存中处理的比率， PGA缓存命中率

W/A MB processes: 指在Workarea中处理过的数据的量，单位为MB

Extra W/A MB Read/Written : 指额外从磁盘上读写的工作区数据,单位为 MB

7-4 PGA Aggr Target Stats

PGA Aggr Target Stats

No data exists for this section of the report.

PGA Aggr Target(M) ：本质上就是pga_aggregate_target

Auto PGA Target(M) : 在自动PGA 管理模式下实际可用的工作区内存 “aggregate PGA auto target “，

W/A PGA Used(M) ：所有的工作区workarea(包括manual和 auto)使用的内存总和量，单位为MB

%PGA W/A Mem: 分配给workarea的内存量占总的PGA的比例，

%Auto W/A Mem : AUTO 自动工作区管理所控制的内存占总的workarea内存的比例

%Man W/A Mem : MANUAL 手动工作区管理所控制的内存占总的workarea内存的比例

Global Mem Bound(K) : 指在自动PGA管理模式下一个工作区所能分配的最大内存(注意一个SQL执行过程中可能有多个工作区workarea)。Global Mem Bound(K)这个指标在实例运行过程中将被持续性的修正，以反应数据库当时工作区的负载情况。显然在有众多活跃工作区的系统负载下相应地Global Mem Bound将会下降。但应当保持global bound值不要小于1 MB ，否则建议调高pga_aggregate_target

7-5 PGA Aggr Target Histogram

PGA Aggr Target Histogram

Optimal Executions are purely in-memory operations

Low Optimal	High Optimal	Total Execs	Optimal Execs
2K	4K	96,633	96,633
64K	128K	504	504
128K	256K	88	88
256K	512K	70	70
512K	1024K	4,616	4,616
1M	2M	2,915	2,915
2M	4M	643	643
4M	8M	6	6
8M	16M	9	9
16M	32M	6	6
32M	64M	2	2

Low Optimal：此行所包含工作区workarea最适合内存要求的下限

High Optimal：此行所包含工作区workarea最适合内存要求的上限

Total Execs: 在 Low Optimal~High Optimal 范围工作区内完成的总执行数

Optimal execs: optimal执行是指完全在PGA内存中完成的执行次数

1-pass Execs : 指操作过程中仅发生1次磁盘读取的执行次数

M-pass Execs: 指操作过程中发生了1次以上的磁盘读取，频发磁盘读取的执行次数

7-6 PGA Memory Advisory

PGA Memory Advisory

When using Auto Memory Mgmt, minimally choose a pga_aggregate_target value where Estd PGA Overalloc Count is 0

PGA Target Est (MB)	Size Factr	W/A MB Processed	Estd Extra W/A MB Read/ Written to Disk	Estd PGA Cache Hit %	Estd PGA Overalloc Count	Estd Time
564	0.13	8,289,305.45	287,052.29	97.00	45	4,922,528,131
1,128	0.25	8,289,305.45	84,121.76	99.00	0	4,806,053,130
2,256	0.50	8,289,305.45	83,858.20	99.00	0	4,805,901,855
3,384	0.75	8,289,305.45	83,858.20	99.00	0	4,805,901,855
4,512	1.00	8,289,305.45	60,642.91	99.00	0	4,792,577,099
5,414	1.20	8,289,305.45	60,521.43	99.00	0	4,792,507,373
6,317	1.40	8,289,305.45	60,521.43	99.00	0	4,792,507,373
7,219	1.60	8,289,305.45	60,521.43	99.00	0	4,792,507,373
8,122	1.80	8,289,305.45	60,521.43	99.00	0	4,792,507,373
9,024	2.00	8,289,305.45	60,521.43	99.00	0	4,792,507,373
13,536	3.00	8,289,305.45	60,521.43	99.00	0	4,792,507,373
18,048	4.00	8,289,305.45	60,521.43	99.00	0	4,792,507,373
27,072	6.00	8,289,305.45	60,521.43	99.00	0	4,792,507,373
36,096	8.00	8,289,305.45	60,521.43	99.00	0	4,792,507,373

PGA Target Est (MB) 用以评估的 PGA_AGGREGATE _TARGET值

Size Factr ，当前用以评估的PGA_AGGREGATE _TARGET和当前实际设置的PGA_AGGREGATE_TARGET之间的比例因子PGA Target Est PGA_AGGREGATE_TARGE

W/A MB Processed ：workarea中要处理的数据量，单位为MB

Estd Extra W/A MB Read/ Written to Disk : 以 one-pass 、M-Pass方式处理的数据量预估值，单位为MB

Estd P Cache Hit % : 预估的PGA缓存命中率

Estd PGA Overalloc Count: 预估的PGA过载量,所述PGA_AGGREGATE _TARGET仅是一个目标值，无法真正限制PGA内存的使用，当出现 PGA内存硬性需求时会产生PGA overallocate 过载(When using Auto Memory Mgmt, minimally choose a pga_aggregate_target value where Estd PGA Overalloc Count is 0)

7-7 Shared Pool Advisory

Shared Pool Advisory

SP: Shared Pool Est LC: Estimated Library Cache Factr: Factor
Note there is often a 1:Many correlation between a single logical object in the Library Cache, and the physical number of memory objects associated with it. Therefore comparing the number of Lib Cache objects (e.g. in v$librarycache), with the number of Lib Cache Memory Objects is invalid.

Shared Pool Size(M)	SP Size Factr	Est LC Size (M)	Est LC Mem Obj	Est LC Time Saved (s)	Est LC Time Saved Factr	Est LC Load Time (s)	Est LC Load Time Factr	Est LC Mem Obj Hits (K)
3,040	0.60	416	60,195	3,762,898	0.44	5,169,739	17.62	974,023
3,552	0.70	927	86,857	5,146,665	0.60	3,785,972	12.90	1,140,581
4,064	0.80	1,438	115,517	6,393,850	0.74	2,538,787	8.65	1,276,670
4,576	0.90	1,949	138,557	7,580,669	0.88	1,351,968	4.61	1,402,383
4,768	0.94	2,141	147,522	8,009,651	0.93	922,986	3.15	1,447,003
4,800	0.94	2,173	149,280	8,079,477	0.94	853,160	2.91	1,454,012
4,832	0.95	2,205	150,603	8,149,115	0.94	783,522	2.67	1,460,993
4,864	0.96	2,237	152,440	8,219,195	0.95	713,442	2.43	1,468,076
4,896	0.96	2,269	153,686	8,287,515	0.96	645,122	2.20	1,474,824
4,928	0.97	2,301	155,511	8,355,821	0.97	576,816	1.97	1,481,547
4,960	0.97	2,333	157,308	8,424,114	0.98	508,523	1.73	1,488,291
4,992	0.98	2,365	158,859	8,492,427	0.98	440,210	1.50	1,495,136
5,024	0.99	2,397	160,249	8,559,773	0.99	372,864	1.27	1,502,818
5,056	0.99	2,429	161,635	8,600,300	1.00	332,337	1.13	1,507,575
5,088	1.00	2,461	163,084	8,639,210	1.00	293,427	1.00	1,512,485
5,120	1.01	2,493	165,151	8,641,318	1.00	291,319	0.99	1,513,085
5,152	1.01	2,523	167,106	8,641,402	1.00	291,235	0.99	1,513,123
5,184	1.02	2,555	168,990	8,641,499	1.00	291,138	0.99	1,513,163
5,216	1.03	2,587	170,484	8,641,600	1.00	291,037	0.99	1,513,203
5,248	1.03	2,619	172,220	8,641,702	1.00	290,935	0.99	1,513,237
5,280	1.04	2,651	173,866	8,641,779	1.00	290,858	0.99	1,513,268
5,312	1.04	2,683	175,560	8,641,852	1.00	290,785	0.99	1,513,300
5,344	1.05	2,715	177,195	8,641,932	1.00	290,705	0.99	1,513,332
5,376	1.06	2,747	178,918	8,642,006	1.00	290,631	0.99	1,513,363
5,600	1.10	2,971	192,344	8,642,523	1.00	290,114	0.99	1,513,598
6,112	1.20	3,483	219,847	8,643,633	1.00	289,004	0.98	1,514,118
6,624	1.30	3,995	248,721	8,644,548	1.00	288,089	0.98	1,514,525
7,136	1.40	4,507	274,955	8,645,425	1.00	287,212	0.98	1,514,932
7,648	1.50	5,019	295,115	8,646,263	1.00	286,374	0.98	1,515,331
8,160	1.60	5,530	323,720	8,647,049	1.00	285,588	0.97	1,515,713
8,672	1.70	6,041	364,001	8,647,797	1.00	284,840	0.97	1,516,028
9,184	1.81	6,552	387,864	8,648,513	1.00	284,124	0.97	1,516,322
9,696	1.91	7,063	415,296	8,649,184	1.00	283,453	0.97	1,516,607
10,208	2.01	7,575	443,210	8,649,818	1.00	282,819	0.96	1,516,874

Shared Pool Size(M) : 用以评估的shared pool共享池大小，在AMM ASMM环境下 shared_pool 大小都可能浮动

SP Size Factr ：共享池大小的比例因子，（Shared Pool Size for Estim SHARED_POOL_SIZE）

Estd LC Size(M) : 评估的 library cache大小,单位为MB，因为是shared pool中包含 library cache 当然还有其他例如row cache

7-8 SGA Target Advisory

SGA Target Advisory

SGA Target Size (M)	SGA Size Factor	Est DB Time (s)	Est Physical Reads
4,192	0.50	5,510,525	3,005,545,869
5,240	0.63	3,074,402	3,005,545,869
6,288	0.75	1,609,103	408,575,030
7,336	0.88	1,558,431	80,417,920
8,384	1.00	1,549,598	23,006,128
9,432	1.13	1,547,429	23,006,128
10,480	1.25	1,545,569	17,321,314
11,528	1.38	1,543,864	17,321,314
12,576	1.50	1,542,315	17,321,314
13,624	1.63	1,540,920	17,321,314
14,672	1.75	1,539,836	14,284,505
15,720	1.88	1,539,526	12,365,794
16,768	2.00	1,539,526	11,896,469

SGA target Size : 用以评估的sga target大小 (sga_target)

SGA Size Factor: SGA Size的比例因子，

Est DB Time (s): 评估对应于该指定sga target size会产生多少量的DB TIME，单位为秒

Est Physical Reads：评估对应该指定的sga target size 会产生多少的物理读

7-9 Streams Pool Advisory

Streams Pool Advisory

Size for Est (MB)	Size Factor	Est Spill Count	Est Spill Time (s)	Est Unspill Count	Est Unspill Time (s)
32	0.50	0	0	0	0
64	1.00	0	0	0	0
96	1.50	0	0	0	0
128	2.00	0	0	0	0
160	2.50	0	0	0	0
192	3.00	0	0	0	0
224	3.50	0	0	0	0
256	4.00	0	0	0	0
288	4.50	0	0	0	0
320	5.00	0	0	0	0
352	5.50	0	0	0	0
384	6.00	0	0	0	0
416	6.50	0	0	0	0
448	7.00	0	0	0	0
480	7.50	0	0	0	0
512	8.00	0	0	0	0
544	8.50	0	0	0	0
576	9.00	0	0	0	0
608	9.50	0	0	0	0
640	10.00	0	0	0	0

Size for Est (MB) : 用以评估的 streams pool大小

Size Factor ：streams pool大小的比例因子

Est Spill Count ：评估出的当使用该大小的流池时 message溢出到磁盘的数量 ESTD_SPILL_COUNT

Est Spill Time (s)：评估出的当使用该大小的流池时 message溢出到磁盘的耗时，单位为秒 ESTD_SPILL_TIME

Est Unspill Count：评估的当使用该大小的流池时　message unspill 即从磁盘上读取的数量 ESTD_UNSPILL_COUNT

Est Unspill Time (s) ：评估的　当使用该大小的流池时 message unspill 即从磁盘上读取的耗时，单位为秒 ESTD_UNSPILL_TIME

8 Wait Statistics

8-1 Buffer Wait Statistics

Buffer Wait Statistics

ordered by wait time desc, waits desc

Class	Waits	Total Wait Time (s)	Avg Time (ms)
data block	26	0	1
undo header	4	0	3

该环节是对缓冲池中各类型(class) 块等待的汇总信息，wait的原因一般是 buffer busy waits 和 read by other session

class 数据块的class，一个oracle数据块即有class 属性还有type 属性，数据块中记录type属性(KCBH)，而在buffer header里存有class属性(X$BH.class)

Waits: 该类型数据块的等待次数

Total Wait Time (s) : 该类型数据块的合计等待时间单位为秒

Avg Time (ms) : 该类型数据块平均每次等待的耗时，单位 ms

对于INSERT ON INDEX 引起的争用：

使用反向索引key
使用HASH分区和本地索引
可能的情况下减少index的density

8-2 Enqueue Activity

Enqueue Activity

only enqueues with waits are shown
Enqueue stats gathered prior to 10g should not be compared with 10g data
ordered by Wait Time desc, Waits desc

Enqueue Type (Request Reason)	Requests	Succ Gets	Waits	Wt Time (s)	Av Wt Time(ms)
KO-Multiple Object Checkpoint (fast object checkpoint)	230	230	23	1	48.70
TX-Transaction (row lock contention)	8	8	8	0	25.00
RO-Multiple Object Reuse (fast object reuse)	30	30	2	0	10.00

Enqueue Type (Request Reason) enqueue 队列的类型，

enqueue type是队列锁所要保护的资源如TM 表锁CF 控制文件锁，

enqueue mode 是持有队列锁的模式 (SS、SX 、S、SSX、X)

Requests : 申请对应的enqueue type资源或者队列转换(enqueue conversion 例如 S 转 SSX ) 的次数

Succ Gets ：对应的enqueue被成功申请或转换的次数
Failed Gets ：对应的enqueue的申请　或者转换失败的次数
Waits　：由对应的enqueue的申请或者转换而造成等待的次数
Wt Time (s) ：由对应的enqueue的申请或者转换而造成等待的等待时间
Av Wt Time(ms) ：由对应的enqueue的申请或者转换而造成等待的平均等待时间， Wt Time (s) Waits ,单位为m

主要的enqueue 等待事件：
enq: TX – row lock/index contention、allocate ITL等待事件
enq: TM – contention等待事件

8-3 Undo Segment Summary

Undo Segment Summary

Min/Max TR (mins) - Min and Max Tuned Retention (minutes)
STO - Snapshot Too Old count, OOS - Out of Space count
Undo segment block stats:
uS - unexpired Stolen, uR - unexpired Released, uU - unexpired reUsed
eS - expired Stolen, eR - expired Released, eU - expired reUsed

Undo TS#	Num Undo Blocks (K)	Number of Transactions	Max Qry Len (s)	Max Tx Concurcy	Min/Max TR (mins)	STO/ OOS	uS/uR/uU/ eS/eR/eU
2	6.49	14,017	905	4	16/26.1	0/0	0/0/0/0/0/0

Undo Extent有三种状态 active 、unexpired 、expired

active => extent中包括了活动的事务，active的undo extent 一般不允许被其他事务重用覆盖

Max Qry Len (s) 该时段内持续最久的查询时间，单位为秒

Max Tx Concy 该时段内最大的事务并发量

10-1 Latch Activity

Latch Activity

"Get Requests", "Pct Get Miss" and "Avg Slps/Miss" are statistics for willing-to-wait latch get requests
"NoWait Requests", "Pct NoWait Miss" are for no-wait latch get requests
"Pct Misses" for both should be very close to 0.0

Latch Name	Get Requests	Pct Get Miss	Avg Slps Miss	NoWait Requests	Pct NoWait Miss
AQ deq hash table latch	1	0.00		0
ASM db client latch	2,458	0.00		0
ASM map operation hash table	1	0.00		0
ASM network state latch	57	0.00		0
AWR Alerted Metric Element list	28,644	0.00		0
Change Notification Hash table latch	1,202	0.00		0
Consistent RBA	11,572	0.00		0
DML lock allocation	40,814	0.00		0
Event Group Locks	559	0.00		0
FAL Queue	144	0.00		0
FOB s.o list latch	617	0.16	0.00	0
File State Object Pool Parent Latch	1	0.00		0
I/O Staticstics latch	1	0.00		0
IPC stats buffer allocation latch	1	0.00		0
In memory undo latch	73,620	0.02	0.67	16,139	0.00
JS Sh mem access	7	0.00		0
JS mem alloc latch	4	0.00		0
JS queue access latch	5	0.00		0
JS queue state obj latch	26,134	0.00		0

latch name Latch：闩的名字

Get Requests latch被以willing-to-wait模式申请并获得的次数

Pct Get Miss miss是指latch被以willing-to-wait 模式申请但是申请者必须等待的次数， Pct Get Miss = Miss/Get Requests ; miss可以从后面的Latch Sleep Breakdown 获得

Avg Slps Miss Sleep 是指latch被以willing-to-wait模式申请最终导致session需要sleep以等待该latch的次数；

Avg Slps Miss = Sleeps/ Misses ; Sleeps可以从后面的Latch Sleep Breakdown 获得
Wait Time (s) 指花费在等待latch上的时间，单位为秒

NoWait Requests 指latch被以no-wait模式来申请的次数

Pct NoWait Miss 以no-wait模式来申请latch但直接失败的次数

对于高并发的latch例如cache buffers chains，其Pct Misses应当十分接近于0
一般的调优原则：
如果latch : cache buffers chains是 Top 5 事件，则需要考虑优化SQL减少全表扫描并减少Top buffer gets SQL语句的逻辑读
如果latch : redo copy 、redo allocation 等待较多，则可以考虑增大LOG_BUFFER

如果latch:library cache 发生较多，则考虑增大shared_pool_size

10-2 Latch Sleep Breakdown

Latch Sleep Breakdown

ordered by misses desc

Latch Name	Get Requests	Misses	Sleeps	Spin Gets
row cache objects	23,008,103	68,048	551	67,505
cache buffers chains	160,393,751	48,378	60	48,326
session idle bit	23,321,979	7,797	542	7,263
shared pool	4,190,965	5,811	878	5,031
session allocation	3,152,314	350	10	340
redo allocation	67,103	22	2	20
mostly latch-free SCN	11,861	21	4	17
In memory undo latch	73,620	18	12	6
simulator hash latch	4,671,494	18	4	14
shared pool simulator	276,285	14	3	11
enqueue hash chains	297,395	10	1	9
call allocation	2,328	3	1	2
simulator lru latch	7,463	3	2	1

latch name Latch闩的名字

Get Requests latch被以willing-to-wait模式申请并获得的次数

misses 是指latch被以willing-to-wait 模式申请但是申请者必须等待的次数

10-3 Latch Miss Sources

Latch Miss Sources

only latches with sleeps are shown
ordered by name, sleeps desc

Latch Name	Where	Sleeps	Waiter Sleeps
In memory undo latch	ktiFlush: child	11	11
In memory undo latch	kticmt: child	3	1
In memory undo latch	ktichg: child	1	0
cache buffers chains	kcbgtcr: fast path (cr pin)	45	11
cache buffers chains	kcbrls_2	18	50
cache buffers chains	kcbchg1: mod cr pin	14	5
cache buffers chains	kcbgtcr: kslbegin excl	13	11
cache buffers chains	kcbgcur_2	5	1
cache buffers chains	kcbgcur_4	1	1
cache buffers chains	kcbgtcr_2	1	0
call allocation	ksudlp: top call	1	0
enqueue hash chains	ksqrcl	1	0
mostly latch-free SCN	kcslcu3	4	4
redo allocation	kcrfw_redo_write: before write	2	0
row cache objects	kqreqd: reget	303	2
row cache objects	kqrpre: find obj	127	427
row cache objects	kqrso	94	93
row cache objects	kqreqd	24	28
session allocation	ksucri_int : SSO	5	0
session allocation	ksuxds	5	10
session idle bit	ksupuc: set busy	396	204
session idle bit	ksupuc: clear busy	133	235
session idle bit	ksuxds	59	149
shared pool	kghalo	566	416
shared pool	kghalp	188	164
shared pool	kghupr1	90	263
shared pool	kgh_heap_sizes	16	12
shared pool	kghasp	10	6
shared pool	kghfre	7	12
shared pool	kghfrunp: clatch: nowait	2	0
shared pool	kghfree_extents: scan	1	5
shared pool	kghfrunp: clatch: wait	1	0
shared pool simulator	kglsim_upd_newhp	3	2
simulator hash latch	kcbsacc: lookup dba	4	4
simulator lru latch	kcbs_simulate: simulate set	2	0

latch name Latch闩的名字

where : 指哪些代码路径内核函数持有过这些该latch ，而不是哪些代码路径要申请这些latch；例如kcbgtcr函数的作用是Get a block for Consistent read，其持有latch :cache buffers chain是很正常的事情

NoWait Misses: 以no-wait模式来申请latch但直接失败的次数

Sleeps: 指latch被以willing-to-wait模式申请最终导致session需要sleep以等待该latch的次数 time of sleeps resulted in making the latch request

Waiter Sleeps：等待者休眠的次数 times of sleeps that waiters did for each where; Sleep 是阻塞者等待的次数， Waiter Sleeps是被阻塞者等待的次数

10-4 Mutex Sleep Summary

Mutex Sleep Summary

ordered by number of sleeps desc

Mutex Type	Location	Sleeps
Cursor Pin	kksfbc [KKSCHLPIN1]	4,526
Cursor Pin	kksLockDelete [KKSCHLPIN6]	2,996
Cursor Pin	kkslce [KKSCHLPIN2]	2,116
Library Cache	kglhdgn2 106	285
Library Cache	kglget2 2	172
Library Cache	kglget1 1	164
Library Cache	kgllkdl1 85	105
Library Cache	kglhdgh1 64	65
Library Cache	kglhdgn1 62	62
Library Cache	kglpndl1 95	57
Library Cache	kgldtin1 42	48
Library Cache	kglrfcl1 79	46
Library Cache	kglpin1 4	37
Library Cache	kgllkc1 57	26
Library Cache	kglobpn1 71	25
Library Cache	kglpnal1 90	15
hash table	kkscsSearchChildList [KKSHBKLOC2]	3
Library Cache	kglllal3 111	1

11 segment statistics

11-1 Segments by Logical Reads

Segments by Logical Reads

Total Logical Reads: 81,939,646
Captured Segments account for 92.6% of Total

Owner	Tablespace Name	Obj. Type	Logical Reads	%Total
O	USERS	TABLE	41,127,232	50.19
O	1	TABLE	2,717,760	3.32
O	1	TABLE	2,600,624	3.17
O	USERS	TABLE	2,581,152	3.15
O	1	TABLE	2,211,792	2.70

owner : 数据段的所有者

Tablespace Name: 数据段所在表空间名

Object Name : 对象名

Subobject Name：子对象名，例如一个分区表的某个分区

obj Type: 对象类型一般为TABLE INDEX 或者分区或子分区

Logical Reads ：该数据段上发生过的逻辑读，单位为块数*次数

%Total : 占总的逻辑读的百分比,当前对象上发生过的逻辑读/Total DB逻辑读

11-2 Segments by Physical Reads

Segments by Physical Reads

Total Physical Reads: 1,427,765
Captured Segments account for 99.2% of Total

Owner	Tablespace Name	Object Name	Obj. Type	Physical Reads	%Total
O			TABLE	783,324	54.86
O	USERS		TABLE	624,402	43.73
O	USERS	$$	LOB	8,316	0.58
O			TABLE	145	0.01
O	USERS		TABLE	128	0.01

Physical Reads: 该数据段上发生过的物理读，单位为块数*次数

%Total : 占总的物理读的百分比，(当前对象上发生过的逻辑读/Total DB 逻辑读)

11-3 Segments by Physical Read Requests

Segments by Physical Read Requests

Total Physical Read Requests: 31,234
Captured Segments account for 64.2% of Total

Owner	Tablespace Name	Object Name	Obj. Type	Phys Read Requests	%Total
O	USERS	$$	LOB	8,195	26.24
O			TABLE	6,180	19.79
O	USERS		TABLE	5,242	16.78
O			TABLE	145	0.46
O	USERS		TABLE	128	0.41

Phys Read Requests ：物理读的申请次数

%Total ：(该段上发生的物理读的申请次数/ physical read IO requests)

12 Dictionary Cache Stats

Dictionary Cache Stats

"Pct Misses" should be very low (< 2% in most cases)
"Final Usage" is the number of cache entries being used

Cache	Get Requests	Pct Miss	Mod Reqs	Final Usage
dc_awr_control	65	0.00	2	1
dc_files	6	100.00	0	6
dc_global_oids	435	4.14	0	308
dc_histogram_data	3,518,709	0.05	0	6,216
dc_histogram_defs	1,307,798	0.28	1	7,437
dc_object_grants	38	7.89	0	26
dc_objects	1,750,326	0.06	48	2,177
dc_profiles	282	0.00	0	1
dc_rollback_segments	886	0.00	0	22
dc_segments	546,372	0.06	2	6,319
dc_sequences	339	13.57	339	39
dc_tablespaces	1,316,800	0.00	0	7
dc_users	1,569,629	0.00	0	125
global database name	2,591	0.00	0	1
outstanding_alerts	30	0.00	0	5
sch_lj_oids	4	25.00	0	1

Get Requests 申请获取该数据字典缓存对象的次数 gets

Miss : GETMISSES 申请获取该数据字典缓存对象但 miss的次数

Pct Miss ：　GETMISSES Gets ， Miss的比例，这个pct miss应当非常低小于2%，否则有出现大量row cache lock的可能

Scan Reqs：扫描申请的次数，

Pct Miss：SCANMISSES/SCANS

13 Library Cache Activity

Library Cache Activity

"Pct Misses" should be very low

Namespace	Get Requests	Pct Miss	Pin Requests	Pct Miss	Reloads	Invali- dations
ACCOUNT_STATUS	318	0.31	0		0	0
BODY	1,578	1.08	3,444	0.64	2	0
CLUSTER	179	1.12	179	1.12	0	0
DBLINK	320	0.63	0		0	0
EDITION	189	0.00	363	0.00	0	0
INDEX	42	4.76	42	33.33	12	0
OBJECT ID	1	100.00	0		0	0
QUEUE	63	0.00	215	0.00	0	0
SCHEMA	639	0.16	0		0	0
SQL AREA	317,164	20.79	2,489,844	8.74	7,179	3,253
SQL AREA BUILD	75,120	96.34	0		0	0
SQL AREA STATS	74,461	95.31	74,461	95.31	0	0
SUBSCRIPTION	7	0.00	7	0.00	0	0
TABLE/PROCEDURE	96,208	0.35	374,862	0.58	1,068	0
TRIGGER	2,027	3.75	11,088	1.25	61	0

NameSpace library cache 的命名空间

GETS Requests 该命名空间所包含对象的library cache lock被申请的次数

14 SGA信息

14 -1 SGA Memory Summary

SGA Memory Summary

SGA regions	Begin Size (Bytes)	End Size (Bytes) (if different)
Database Buffers	3,019,898,880
Fixed Size	2,265,984
Redo Buffers	4,460,544
Variable Size	10,435,431,552

14-2 SGA breakdown difference

SGA breakdown difference

ordered by Pool, Name
N/A value for Begin MB or End MB indicates the size of that Pool/Name was insignificant, or zero in that snapshot

Pool	Name	Begin MB	End MB	% Diff
java	free memory	128.00	128.00	0.00
large	PX msg pool	7.81	7.81	0.00
large	free memory	152.19	152.19	0.00
shared	KGLDA	68.12	71.25	4.59
shared	KGLH0	1,322.65	1,374.36	3.91
shared	KGLHD	201.44	206.36	2.44
shared	SQLA	1,710.04	1,874.21	9.60
shared	free memory	1,291.59	1,080.89	-16.31
shared	kglsim heap	69.35	69.35	0.00
shared	kglsim object batch	119.14	119.14	0.00
streams	free memory	63.94	63.94	0.00
	buffer_cache	2,880.00	2,880.00	0.00
	fixed_sga	2.16	2.16	0.00
	log_buffer	4.25	4.25	0.00

Pool 内存池的名字

Name 内存池中细分组件的名字例如KGLH0 存放KEL Heap 0 、SQLA存放SQL执行计划等

Begin MB 快照开始时该组件的内存大小
End MB 快照结束时该组件的内存大小
% Diff 差异百分比

以上

oracle

文章转载自数据与人，如果涉嫌侵权，请发送邮件至：contact@modb.pro进行举报，并提供相关证据，一经查实，墨天轮将立刻删除相关内容。

绝对干货！|【性能调优】Oracle AWR报告指标全解析

WORKLOAD REPOSITORY report for

Time Model Statistics

Time Model Statistics几个特别有用的时间指标：

Foreground Wait Class

Foreground Wait Events

Background Wait Events

Operating System Statistics

Service Statistics

Service Wait Class Stats

User I/O Total Wts : 对应该服务名下用户I/O类等待的总的次数

SQL ordered by Elapsed Time

SQL ordered by CPU Time

SQL ordered by Reads

SQL ordered by Physical Reads (UnOptimized)

SQL ordered by Executions

SQL ordered by Parse Calls

SQL ordered by Version Count

Other Instance Activity Stats

Instance Activity Stats - Absolute Values

Tablespace IO Stats

File IO Stats

Buffer Pool Statistics

Checkpoint Activity

Instance Recovery Stats

Buffer Pool Advisory

PGA Aggr Summary

PGA Aggr Target Histogram

Shared Pool Advisory

SGA Target Advisory

Streams Pool Advisory

Buffer Wait Statistics

Enqueue Activity

Undo Segment Summary

Undo Extent有三种状态 active 、unexpired 、expired

Latch Activity

Latch Sleep Breakdown

Mutex Sleep Summary

Segments by Logical Reads

Segments by Physical Reads

Segments by Physical Read Requests

Dictionary Cache Stats

Library Cache Activity

SGA Memory Summary

SGA breakdown difference

评论