Friday, December 12, 2014

Guarding Against Brain Dumps

I have been trying unsuccessfully for some time to get my Oracle Certification Prep site listed as being safe on the CertGuard website. However, I have never succeeded in getting a response back from them.  Until recently I simply left it at that.  However, a couple of days ago I did some more searching and found several signs that the site has been stagnant since around 2011.

I asked a colleague who writes about certifications across multiple vendors, Ed Tittle, if he knew anything about the issue.  Ed did some digging and created an article at GoCertify.  I'm not going to recreate here what he already wrote.  If you are interested in avoiding being caught out by brain dump sites, you should check out his article.

Friday, November 28, 2014

Oracle 12c Performance Certification Released to Beta

The first of the new expert exams that I wrote about back in May has just entered beta.  1Z1-064: Oracle Database 12c: Performance Management and Tuning is the 12c update to the legacy exam 1Z0-054: Oracle Database 11g: Performance Tuning. The beta is currently scheduled to run until late February. Betas are often extended if not enough candidates take the exam by the original deadline to provide a good statistical sample. However, I expect this exam will be popular enough that the exam may close on its original deadline.

I compared the topics of 1Z0-064 with 1Z0-054. Not surprisingly, the two lists are closely related, with the 12c exam being largely a refinement of the 11g incarnation.  Some of the difference between the two exams include:

Added topics:

  • Implement Real-Time Database Operation Monitoring
  • Understand and configure the Database Resource Manager

Removed Topics:

  • Identify performance issues & set tuning priorities
  • Interpret tuning diagnostics  
  • Tune for life cycle phase 
  • Using Statspack
  • Use Enterprise Manager Monitoring  
  • Identify the key tuning components of the alert logs 

I find it interesting that Statspack was removed. Obviously AWR is Oracle's preferred performance tuning solution. However, unless that stops being a separately licensed option, a significant number of Oracle DBAs will continue to use Statspack for diagnosing performance problems.  Ignoring this popular tuning option comes across as petty to me.

The 12c exam has considerably fewer topics than the 11g version (32 vs 54). Much of this comes in the form of topic consolidation. In several instances there are subjects that were split into several separate topics in 1Z0-054 that have been rolled into a single topic in 1Z0-064. Some examples include:

  • Configure and manage services 
  • Use services with client applications, Database Resource Manager and Scheduler 
  • Configure services aggregation, tracing, and set performance-metric thresholds

Rolled into:

  • Configure and use services to monitor database performance

  • Describe the buffer cache architecture
  • Explain the symptoms that indicate a buffer cache problem   
  • Tune the buffer cache for performance issues 
Rolled into:
  • Diagnose and resolve performance issues related to the buffer cache
I consider it unlikely that the information being tested on these consolidations has changed significantly. Most likely the questions have all simply been consolidated into a large bucket.

Performance is always a major factor for enterprise databases. No organization ever objects to their applications running faster or data being available more rapidly. This certification provides one more way for Oracle Database administrators to expand on their skill set.

Saturday, November 15, 2014

Finally! The revised version of my 1Z0-047 Study Guide is complete and published.

When I found out about the changes being made to the 1Z0-047 exam (and reported on it in this blog), I pulled up the master document for my SQL Expert study guide and outlined the changes that needed to be made in order for it to be comprehensive enough for the new exam. It really should not have taken me this long to complete the revisions, but it did. I had originally hoped to get an updated book into print before the revised topics even made it to production. That plan went out the window two months ago...

Be that as it may, the guide is done now.  Part of what took so long was that the deeper I got into the revision process, the further I went beyond my original intent. Ultimately, a hefty percentage of the guide ended up being revised and expanded. This was one of the earliest study guides in the Oracle Certification Prep series. Since creating it, I have gotten much better at the process. In addition, I have created a standard set of formatting rules used in creating books in the series -- which this guide did not follow at all.

While this project took longer than I planned, I am very happy with the result. The organization is cleaner. I have added more examples and reworded some paragraphs that I felt were not clear enough in the original. The guide published this afternoon is 40% larger than the original and none of the new content is 'fluff'. I hope that anyone who purchases the revised guide will find that it is a valuable resource in preparing for the exam.

Saturday, September 13, 2014

Causality and Oracle Expertise

I was involved in a discussion on LinkedIn about the proper timing for an Oracle PL/SQL developer to take the 1Z0-146: Oracle Database 11g: Advanced PL/SQL exam. One of the posters was indicating that while it was appropriate for people new to PL/SQL to study for and take the earlier exam: 1Z0-144: Oracle Database 11g: Program with PL/SQL, they should wait until they have some job experience before taking the 146 exam to earn the Oracle Advanced PL/SQL Developer Certified Professional credential. My own opinion is that waiting until you have a few years of experience is not a bad idea, but it is certainly not something that is necessary for everyone.

The difference between our positions lies in how I view Oracle certifications, their relationship to Oracle knowledge/expertise and (as the title suggests) the causality between the two. Causality is an uncommon word -- even for native English speakers (unless you happen to be a theoretical physicist). Basically it refers to cause and effect. The 'effect' that I am referring to this case is getting employers to recognize you as being knowledgeable about Oracle PL/SQL. There is a widespread belief that earning a given certification will cause employers to view you as being skilled with Oracle.

It does not work that way in the real world. I have interviewed people for Oracle positions many times over the years. I can tell you that I do not automatically assume that because someone has earned a certification that they have all of the relevant skills. A certification means that someone may have a certain set of knowledge. During the interview, I will generally ask questions related to their certifications to get a feel for whether they do or not. If a candidate has the Advanced PL/SQL certification and their resume shows a previous job developing PL/SQL, I will be more likely to believe that they are knowledgeable of the language. I will still ask the candidate several questions to verify that to my satisfaction, however.

Employers cannot trust certifications by themselves because there are too many people who focus on the piece of paper. Their goal is passing the exam rather than learning the information being tested. In researching my study guides I run across brain dump sites all the time. Even people who do not cheat with brain dumps may simply memorize enough facts to pass the exam without actually understanding the material being tested. To be fair, I do not completely trust job experience either. It is possible to be PL/SQL developer for several years and still be really bad at it. I know this is true -- I have worked with some of them.

So... if certifications are not the cause for employers viewing an individual as being knowledgeable about Oracle PL/SQL, then what is? Well -- not to put to fine a point on it, but the cause is becoming knowledgeable about Oracle PL/SQL and being able to demonstrate that knowledge. Candidates should take the time to really learn all of the information being tested in the 1Z0-144 and 1Z0-146 exams. Practice writing PL/SQL blocks. Read tutorials and articles from some of the masters like Steven Feuerstein.This will give you a core of knowledge about the language even if you have no job experience using it. In an interview, you will be able to demonstrate this knowledge when asked questions and that will generate the desired effect.

My approach to certifications is therefore that they should be used as a road map for becoming more knowledgeable. The focus is not on 'passing the test' but rather on 'learning the information'. Focusing on the test is what leads to brain dumps and other cheats. This is counterproductive. If you get hired on the basis of knowledge you do not possess, it will rapidly become obvious. After you pass the exams, keep reading and practicing and learning. In the IT industry, the more you know, the more valuable you are. As a fresher, that knowledge will help you to sit an interview with more confidence that whatever question is asked, you will be able to answer it. Your increased confidence and your knowledge will both do much more to land the job than any piece of paper.

Wednesday, September 3, 2014

1Z0-047 topics being updated on 15-SEP-14. Don't get caught unawares.

A little less than two weeks from the date I'm writing this post, the topics list for the SQL Expert exam, 1Z0-047, will include a number of new subject areas. Simultaneously, two of the legacy topics will be removed, but their impact to test-takers is minimal. I wrote an article discussing the changes in more detail that was just published at CertMag here.

If you have been preparing for this exam and have purchased any existing study materials such as the Oracle Press OCA Oracle Database SQL Certified Expert Exam Guide by Steve O'Hearn or my Study Guide for 1Z0-047: Oracle Database SQL Expert, they will not cover these topics. I am working on a major update to my guide at this time that should be released in a few weeks. I do not have any information about when (or if) Steve's Oracle Press book will be updated to include the new material.  The full list of topics that are being added is:

  • Using pattern matching to recognize patterns across multiple rows in a table
  • Using the PIVOT and UNPIVOT clause
  • Using the SQL row limiting clause
  • Analytical functions including PERCENTILE_CONT, STDDEV, LAG, LEAD
  • Using 12c enhancements to the DEFAULT clause, invisible columns, virtual columns and identity columns
  • Using explicit default values in INSERT and UPDATE statements
  • Using the new index capabilities, such as invisible indexes and multiple indexes on the same columns.
  • Recursively truncate child tables
  • Create a lateral inline view in a query
  • Using the cross_outer_apply_clause

I have updated the 1Z0-047 Exam Detail page on my website to include links to articles for all of the new topics. As so often happens, many of the new links are from +Tim Hall and point to articles on his enormously informative Oracle-Base website. If you are in the final stages of preparing for the SQL Expert exam, you can use these articles along with the Oracle documentation to learn about the additions.

Monday, June 30, 2014

A Lifeline for 9i & 10g OCAs -- Oracle Releases Two OCA-to-OCP Upgrade Options

Historically, Oracle Certified Associates in the Administration track for one release of the database have had only two options for earning the Oracle Certified Professional designation in a later release:
  1. Complete the OCP credential in the release they hold the OCA in. Then take one or more upgrade exams required to bump their OCP to a more current release.
  2. Start from (almost) scratch to earn the OCA and then the OCP in the later release.

Both options require a minimum of two exams. For people that hold the OCA in 9i, the only option since that track was retired last year has been to start over with the OCA in a later release. The introduction of two new exams by the Oracle Certification team has changed the game. Either exam will allow Oracle Certified Associates to upgrade directly to the 11g or 12c Oracle Certified Professional credential. The beta period for the two exams started June 7th:
  • 1Z0-034: Upgrade Oracle9i/10g OCA to Oracle Database 11g OCP
  • 1Z0-067: Upgrade Oracle9i/10g/11g OCA to Oracle Database 12c OCP

I had speculated in a previous blog post that an exam of this type was on the horizon.  I will freely admit that I did not expect OU to create two exams.  I envisioned a single hyped-up version of the 1Z0-060 exam to upgrade candidates to 12c.  I have been developing study guides for these exams since learning of them. After going through their topic lists, I have to wonder if the Oracle Certification team has a recent new hire named Victor Frankenstein, PhD.  1Z0-034 and 1Z0-067 have been pieced together from a number of earlier exams and both are monsters.

While I had not foreseen that OU would create this exam, the topics list matches what I would have expected... had I been expecting anything at all. The test mixes and matches topics from the 11g Admin II (1Z0-053), Admin I (1Z0-052),  and New Features (1Z0-050) exams as well as some from the 10g Admin II (1Z0-043) and New Features (1Z0-040) exams. In all, this exam has 112 topics in 25 sections.  I was stunned by the sheer scope of the exam, until I realized how much 034 is eclipsed by its big brother.

I am reasonably certain that this exam is the largest by topic count of any that I have ever seen from Oracle University.  The test contains over 220 topics under 42 different sections. The topic list contains most everything that is in the 1Z0-034 exam plus a heaping helping from the three 12c administration exams (1Z0-060, 1Z0-062, 1Z0-063).  Oracle professionals who are upgrading from a 9i OCA in particular will be introduced to a wide array of new features.  Candidates preparing for this test need to plan on setting aside a significant amount of study time in order to have a reasonable chance of passing on the first attempt.

As I assumed in my earlier post, neither of these exams eliminates the hands-on training requirement to earn the OCP designation. The new exams will not eliminate that expense, but they can save candidates the expense of at least one exam. In addition, even though both are heavyweights, preparing for a single test will consume less time than preparing for two or more separate ones. If you decide to pursue one of these exams, make sure that you understand from the outset that while they represent a simple way to upgrade your credentials, that is not at all the same thing as an easy way.  You will need to put in some work for these.

Despite their length, I expect that these exams are likely to be fairly popular.  Certainly I see posts by OCAs fairly regularly asking how they can upgrade their credential to a current release of Oracle.

For people who are interested, I have started gathering links to certification-safe study materials on my website for these exams:
1Z0-034: Upgrade Oracle9i/10g OCA to Oracle Database 11g OCP
1Z0-067: Upgrade Oracle9i/10g/11g OCA to Oracle Database 12c OCP

Monday, June 16, 2014

End-of-Chapter Questions and Oracle Certification Preparation

An Oracle exam candidate sent me the following question recently after reading my study guide for the 1Z0-117: Oracle Database 11g Release 2: SQL Tuning exam:

"One thing which I felt is that there should be mock Q&A's or Quiz to test the knowledge gained after finishing each chapter. Having these gives confidence and a practice before going for the exam."

This is a subject that I have a strong opinion about. I thought it deserved a detailed response rather than a brief email. I will state up front that my study guides will never have practice questions in them for several reasons.

The first problem with adding end-of-chapter questions is economic.  My guides present the information on exam topics in a very concise fashion.  Adding questions that adequately cover the same subjects would easily increase the size of my guides by 50-75%. In addition, creating realistic questions and answers is extremely time consuming. Adding these to my guides might well double the amount of my time required to create them. Either the price of the guides would need to increase significantly or I would need to take a reduced royalty payment (from the increased page count) while simultaneously putting in more hours to create each guide. I do not believe that most of my readers would consider the practice questions to be worth a significant price hike. I know that putting in that much more time for less return would make me considerably less interested in spending my time creating them.

The second issue is that I am not a proponent of studying for exams by answering practice questions. The goal of preparing for Oracle certification exams is to gain as much knowledge about the tested topics as possible. Focusing on learning the answers to specific questions is a terrible habit to fall into. For example, imagine a chapter in a certification preparation book that has five end-of-chapter questions. On taking the self test, someone misses four out of five. This strongly implies that they did not understand the material and should re-read the whole chapter more carefully.  However, many candidates will simply look up the sections in the chapter that dealt with the four questions they missed. Once they have memorized all five answers, they will be under the impression that they are fully prepared for the real test. In reality, it is unlikely they fully understand much of the information outside what the questions covered. If the actual test questions are much different from the practice ones, they are likely to get them wrong.

Finally, the largest problem that I have is that end-of-chapter exercises cannot be close enough to the real test to provide candidates a reasonable measure of how prepared they are for the actual exam. There are several reasons for this, including the following:

  • Generally people answer the questions right after reading the material. This does not mean they will recall it days later when sitting for the exam.
  • Often (as noted above), the exam candidate will check their information in the chapter before answering.  The real test is not open book.
  • The questions are grouped together by topic (i.e. what the chapter was about). This makes answering them easier. The real exam jumps around topics randomly.
  • End of chapter questions have no time constraint.
  • There is absolutely no pressure involved. Test takers are not worried about failing, or whether they really put down the right letter on that one really hard question they answered a few minutes ago.

In my opinion, the primary benefit that practice tests provide is to help exam candidates determine if they are ready for the real test. This is the 'confidence' spoken of in the original question. For that confidence to have any basis in reality, the environment of the self test must be as close as possible to the real exam. It should have the same number of questions and the same time limit as the real test. Exam candidates should treat the practice just like the real exam and not peek in the book or use the internet to help on questions they do not know. I would even recommend taking it at the same time of day and at a terminal close to what will be in the testing center. If all of that is done, the results of a well-written practice exam may be a reasonable approximation of what candidates will get on the actual test.

What all of this adds up to is that I do not believe that end-of chapter exercises provide sufficient value to candidates to be worth the considerable effort that would go into adding them. This is not to say they have zero value. End-of-chapter questions make more sense for the Oracle Press series. The chapters in those books are considerably longer than those in my guides. It is possible -- especially late at night -- to read a complete chapter without actually absorbing the information. If you miss a majority of the practice questions and use that feedback to recognize you should re-read the chapter (perhaps the next day), then they can add value to your study process.