Errata

Optimizing Oracle Performance

Errata for Optimizing Oracle Performance

Submit your own errata for this product.

The errata list is a list of errors and their corrections that were found after the product was released. If the error was corrected in a later version or reprint the date of the correction will be displayed in the column titled "Date Corrected".

The following errata were submitted by our customers and approved as valid errors by the author or editor.

Color key: Serious technical mistake Minor technical mistake Language or formatting error Typo Question Note Update

Version Location Description Submitted By Date submitted Date corrected
Printed
Page xv
first paragraph

the URL to Jared Stills site was incorrect, it NOW READS:
http://www.cybcon.com/~jkstill/util/util_master.html

Anonymous    Dec 01, 2003
Printed
Page xxi
paragraph just following list item #5

Paragraph reads:

We've designed the exercises as a way to get you to try things out in your own
environment. Generally, there are no single, correct answers. The results from a
given exercise will be specific to your systems.

Anonymous    Dec 01, 2006
Printed
Page 32
2nd paragraph

Method R performs reliability in virtually every performance problem situation imaginable;

NOW READS:
Method R performs reliably in virtually every performance problem situation imaginable;

Anonymous    Dec 01, 2003
Printed
Page 79
lid items

Tests in Oracle version 9 show that regardless of whether the execution was done via definer or
invoker rights, the uid and lid always seem to show up in the trace data as the same value: the id
of the user who was logged in and running the procedure.

Anonymous   
Printed
Page 83

After the existing "value" field paragraph, added the following paragraph:

When no value field is emitted to the trace file, it is an indication that the NULL value has been
bound into the placeholder variable. For example, in the fnd_profile.get_specific sample shown
earlier, the absence of a value field for bind 4 indicates that the application has provided the NULL
value for the :val placeholder. The bind value length specification of zero (avl=00) is corroborative
evidence.

Anonymous    Jul 01, 2008
Other Digital Version
83

After the existing "value" field paragraph, added the following paragraph:

When no value field is emitted to the trace file, it is an indication that the NULL value has been
bound into the placeholder variable. For example, in the fnd_profile.get_specific sample shown
earlier, the absence of a value field for bind 4 indicates that the application has provided the NULL
value for the :val placeholder. The bind value length specification of zero (avl=00) is corroborative
evidence.

Anonymous    Jul 01, 2008
Printed
Page 84
Definition of 'w' field inside the 'op' field definition

NOW READS:
w - Number of Oracle blocks written with OS write calls.

Anonymous    Dec 01, 2003
Printed
Page 96
Example 5-7, example input file content

Line break positions HAVE BEEN REVISED, as follows:
# WAIT #2: ...
# WAIT #2: ...
# FETCH #2:...

Anonymous    Jan 01, 2004
Printed
Page 113
Last paragraph

The second occurrence of "TIMED_STATISTICS" in the last paragraph was spelled incorrectly:
TIMED_STATSITICS

NOW READS:
TIMED_STATISTICS

Anonymous    Dec 01, 2003
Printed
Page 117
final code segment before the bear trap icon

select ssession, serial#
from v$session
...

NOW READS:
select sid, serial#
from v$session
...

120, Table 6-2:
MERKUR_S7_FG_ORACLE_013.trc 8.1.7.0.0 OpenVMS 2-1

NOW READS:
MERKUR_S7_FG_ORACLE_013.trc 8.1.7.0.0 OpenVMS 7.2-1

Anonymous    Jan 01, 2004
Printed
Page 118
In the text beside the trap icon

"Do not use DBMS_SYSTEM.START_SQL_TRACE_IN_SESSION..."

NOW READS:
"Do not use DBMS_SYSTEM.SET_SQL_TRACE_IN_SESSION..."

Anonymous    Jul 01, 2005
Printed
Page 133
8th line from the bottom of page, 1st line of code on page

"alter system set events..."

NOW READS:
"alter session set events..."

Anonymous    Jul 01, 2005
Printed
Page 134
Example 6-10, seventh line from the bottom

$frac = sprintf "%6.0f", $frac/(10**($1-6));

NOW READS:
$frac = sprintf "%06.0f", $frac/(10**($1-6));

Anonymous    Dec 01, 2003
Printed
Page 136
Paragraph following the trace excerpt

The duration of the last wait event in the trace NOW READS 4 microseconds, not 4 milliseconds.

Anonymous    Dec 01, 2003
Printed
Page 137
3rd paragraph

"In the Hotsos Sparky data collector, we execute a query that is similar to this one immediately
prior to executing the commands to activate tracing."

NOW READS:
"In the Hotsos Sparky data collector, we execute a query that is similar to this one immediately
prior to executing the commands to activate and deactivate tracing."

Anonymous    Jul 01, 2004
Printed
Page 149
Next-to-final sentence before example 7-3

NOW READS:
The difference between the two marks (c0 and c1) is the approximate amount of CPU
capacity that was consumed by the database call.

Anonymous    Jul 01, 2004
Printed
Page 154
First paragraph

"... before retuning the process to user running state."

NOW READS:
"... before returning the process to user running state."

Anonymous    Jul 01, 2005
Printed
Page 157
1st paragraph

NOW READS: Heisenberg uncertainty principle. (2 instances)

Anonymous    Dec 01, 2003
Printed
Page 157
Inequality above Table 7-1 and Table 7-1 itself

NOW READS:
2.000 0 cs <= e(sub a) <= 2.999 9 cs

since the gettimeofday() function truncates four digits.

Likewise in Table 7-1, the precision, in either zeroes or nines,
NOW READS four decimal places instead of six.

Anonymous    Jan 01, 2004
Printed
Page 158
3rd paragraph

A percentage of quantization error should be 2.1 percent instead of 1.7 percent of the 5.875-tick actual duration.

Note from the Author or Editor:
The submitter is correct: the percentage of quantization error should be 2.1. p = 0.125 / 5.875 * 100% = 2.1276%.

BunditJ  Aug 22, 2014  Dec 04, 2015
Printed
Page 162
In the paragraph following "Polling (sampling)"

"...whatever state the processes is..."

NOW READS:
"...whatever state the process is..."

Anonymous    Jul 01, 2005
Printed
Page 168-169

All (seven) occurrences of the string "255" HAVE BEEN CHANGED to "266", and
a zero HAS BEEN ADDED in the final number of the sidebar at the top of page 169,
so that it NOW READS: 100,000,000,000,000,000,000,000,000.

Anonymous    Jul 01, 2004
Printed
Page 169
Sidebar on "one chance in ten to some large power"

The sidebar mentioned an estimate of 10**80 atoms in the universe, and cites a
siencenet.org.uk URL as one of the sources for that estimate. That URL no longer
functions.

The URLs leading to additional information NOW READ:
http://www.sunspot.noao.edu/sunspot/pr/answerbook/universe.html#q70

http://www.nature.com/nsu/020527/020527-16.html

Anonymous    Jan 01, 2004
Printed
Page 173
2nd Para (last para of 'Process States and Transitions Revisited')

The phrase "as you shall see in one of the case studies in Chapter 12" HAS BEEN DELETED from the last sentence.

Anonymous    Jan 01, 2004
Printed
Page 173
Section Un-Instrumented Oracle Kernel Code, the 2nd paragraph

On page 173 in the section Un-Instrumented Oracle Kernel Code, the 2nd paragraph
reads: u201cChapter 12 lists the number of code segments that are instrumented...u201d

Does this refer to Table 11-1 on page 306 and should therefore read: u201cChapter 11...?u201d

Reader is correct. The right reference now reads Table 11-1.

Anonymous    Jul 01, 2008
Other Digital Version
173
Section Un-Instrumented Oracle Kernel Code, the 2nd paragraph

On page 173 in the section Un-Instrumented Oracle Kernel Code, the 2nd paragraph
reads: u201cChapter 12 lists the number of code segments that are instrumented...u201d

Does this refer to Table 11-1 on page 306 and should therefore read: u201cChapter 11...?u201d

Reader is correct. The right reference now reads Table 11-1.

Anonymous    Jul 01, 2008
Printed
Page 183
"Susceptibility to Overflow and Other Errors" section, second sentence

The sentence NOW READS:
"The problem is that an n-bit counter variable can store only 2^n distinct values."

Anonymous    Dec 01, 2003
Printed
Page 203
Example 8-3 "$Header" line

Program name NOW READS vprof.pl.

Anonymous    Dec 01, 2003
Printed
Page 215
First paragraph

A big part of the problem is a principal that I can best....

NOW READS:
A big part of the problem is a principle that I can best....

Anonymous    Dec 01, 2003
Printed
Page 231
Figure 9-6

The label "Previous arrival" should say "Next arrival"

Anonymous   
Printed
Page 231
1st paragraph, 4th line

"... service time of s time units ..."

NOW READS:
"... service time of S time units ..."

Anonymous    Jul 01, 2004
Printed
Page 238
Figure 9-2

Since P(A=x) is a discrete random variable (not continuous), shouldn't Figure 9-8 be
drawn as a histogram, with a box of unit width centered over each value of x?

"Probability Models for Computer Science" by Sheldon Ross (Harcourt) defines
probability density function only on continuous random variables. For f(x) = P(X =
x) in the second paragraph of pg. 237, he defines that as the probability mass
function.

For a solution to the problem described above, please see the errata listing for p238 at -
http://www.hotsos.com/e-library/oop.html for details.

Anonymous   
Printed
Page 238
3rd paragraph, 7th line

"... a CPUs service rate ..."

NOW READS:
"... a CPU's service rate ..."

Anonymous    Jul 01, 2004
Printed
Page 268
First full paragraph

Similarly, you must ensue that service times are...

NOW READS:
Similarly, you must ensure that service times are...

Anonymous    Dec 01, 2003
Printed
Page 288
First paragraph

"... (the one corresponding to the resource profile in Example 5-2,...)"

NOW READS:
"... (the one corresponding to the resource profile in Example 10-2,...)"

Anonymous    Jul 01, 2004
Printed
Page 290
Third paragraph

"... we were actually taught principals ...."

NOW READS:
"... we were actually taught principles ...."

Anonymous    Dec 01, 2003
Printed
Page 309
3rd paragraph from the top

In "Large c value for a parse call":
If you have many small c values distributes across...

NOW READS:
If you have many small c values distributed across...

Anonymous    Dec 01, 2003
Printed
Page 310
3rd line from bottom

"... patch your Oracle kernel so that the your user action's code ..."

'the' HAS BEEN DELETED from the sentence.

Anonymous    Jul 01, 2004
Printed
Page 339
4th paragraph, line 2

"... we starting looking through ..."

NOW READS:
"... we started looking through ..."

Anonymous    Jul 01, 2004
Printed
Page 340
footnote

principal

NOW READS:
principle

Anonymous    Jul 01, 2004
Printed
Page 343
First para.

"Example 10-3"

NOW READS:
"Example 11-2 on page 322".

Anonymous    Jan 01, 2004
Printed
Page 358
"Methodology" entry

The word "been" was previosuly missing.
Furthermore, the the quoted matter IS NOW offset from the surrounding text in a
manner similar to that shown in the Oracle connection entry on page 359:

the word "methodology" has used as a pretentious...

NOW READS:
the word "methodology" has been used as a pretentious...

Anonymous    Dec 01, 2003
Printed
Page 384
The following index entries for pages 381-388 have been added to the index

buffer busy wait events, 54, 317

Central limit theorem, 169

Greek alphabet, motive for use, 228

Hotsos Profiler, xv, 61

Kyte, Tom, 55, 183, 194-196
Laplace, Pierre Simon de, 169

_SPIN_COUNT, 320b
Still, Jared, xv

tim.pl, 134

tkprof, xv, 61, 320

Trace Analyzer, 61

trcsummary, 61

Anonymous    Jul 01, 2008
Printed
Page 387

START-SQL_TRACE-IN_SESSION events, 118

NOW READS:
SET_SQL_TRACE_IN_SESSION, 118

Anonymous    Jul 01, 2005