DBA · EBS · Intermediate
Oracle Apps DBA: Keeping E-Business Suite Alive
Nineteen modules, nineteen labs, and a system you could hand to somebody else
Nobody notices a DBA until something is down, and by then the decisions that mattered were taken months earlier — when the backup was configured and never restored, when the patch was applied without a rollback, when the test environment was refreshed with real salaries in it. This course is about those decisions. It covers Oracle E-Business Suite and the database underneath it: the architecture, the AD utilities, cloning, patching, concurrent processing, performance, and the security of a system that holds a company's payroll. It assumes you can use a shell and read SQL. It does not assume you have administered a database before.
EGP 3,000
184 lessons · 456 min
What you will be able to do
- Say what an Apps DBA is responsible for, and what changes the day an estate moves to Fusion.
- Read the architecture of a running EBS instance and name every process you find.
- Take a backup you have actually restored, and know how long the restore takes before you need it.
- Clone an environment, and know what a clone does not copy.
- Run a patching cycle with a rollback that exists rather than one you hope for.
- Diagnose a slow system by measuring rather than by tuning what you already suspect.
- Keep a test environment usable without putting real payroll in it.
- Hand a system to somebody else with documentation they can act on.
What you should know first
- You can move around a Linux shell and read a log file without being talked through it.
- You can read SQL. You do not need to have written PL/SQL or administered a database.
- An EBS environment you are allowed to break. A personal or training instance is ideal; a client production system is not.
- No prior DBA experience is assumed. This is not an Oracle certification course and does not follow its syllabus.
What you will cover
1. Module 0 · Orientation
What this course covers, what it deliberately does not, and how to set up an environment you are allowed to destroy.
- 1What you will be able to do3 min
- 2What this course is not2 min
- 3An environment you may break3 min
- 4How to use the labs3 min
2. Module 1 · What an Apps DBA does, and what Fusion took away
The boundary of this course, drawn in the first hour rather than discovered in module 9. If your estate is entirely Fusion SaaS, this says so.
- 1The job, described honestly2 min
- 2Apps DBA versus core DBA3 min
- 3Where the DBA role went in FusionFree preview3 min
- 4What transfers, and what does not3 min
- 5On-premises, IaaS, and the middle ground2 min
- 6Who else touches this system2 min
- 7What breaks: the estate nobody owned3 min
- 8Lab briefing · Describe an estate you have not seen2 min
3. Module 2 · The stack you are responsible for
Every process, file and port in a running EBS instance, named. Almost every diagnostic in this course starts with knowing what should be there.
- 1The two tiers, and why the split matters2 min
- 2What is running right now3 min
- 3The file system layout2 min
- 4Environment files, and sourcing the right one2 min
- 5The context file2 min
- 6AutoConfig, and what it overwrites3 min
- 7Ports, and the ones you will be asked about3 min
- 8Logs: where they are and which one to open first2 min
- 9What breaks: three architecture surprises2 min
- 10Lab briefing · Map a running instance3 min
4. Module 3 · The database itself
For people who did not come from a database background. Enough to reason about what is happening, without pretending to be a course on database internals.
- 1Instance and database: two different things2 min
- 2Starting and stopping, properly3 min
- 3Tablespaces and datafiles3 min
- 4Redo: the reason recovery is possible2 min
- 5Archive log mode, and the day it was off3 min
- 6Undo, and read consistency3 min
- 7Parameters, and the ones that matter here2 min
- 8The data dictionary, as a diagnostic tool2 min
- 9Reading an error you have never seenFree preview2 min
- 10What breaks: four database-tier failures3 min
- 11Lab briefing · Stop it, start it, break it3 min
5. Module 4 · The application tier and the AD utilities
The tools you will spend most of your time in, and the conventions that make them predictable rather than frightening.
- 1What the application tier does2 min
- 2Starting and stopping the services2 min
- 3adadmin, and what it is for2 min
- 4Compiling and generating2 min
- 5adctrl and the concurrent managers2 min
- 6Reading an AD log2 min
- 7Worker processes, and the one that failed3 min
- 8Custom top, and where your code belongs3 min
- 9What breaks: three application-tier failures3 min
- 10Lab briefing · Operate the tier by hand2 min
6. Module 5 · Backup and recovery
FLAGSHIP. The only module where a mistake is unrecoverable in the literal sense. A backup nobody has restored is a belief, not a backup.
- 1What you are actually protecting against3 min
- 2The restore you have never testedFree preview3 min
- 3Recovery objectives, in plain terms2 min
- 4Cold backups: when they are still right2 min
- 5RMAN, from nothing2 min
- 6Full, incremental and the strategy behind them2 min
- 7Backing up the application tier3 min
- 8Point-in-time recovery2 min
- 9Restoring to a different host2 min
- 10How long does a restore actually take2 min
- 11Where the backups live2 min
- 12What breaks: five backup failures2 min
- 13Lab briefing · Lose the database and get it back2 min
7. Module 6 · Cloning
The most-requested DBA task, and the one where the interesting question is what a clone does NOT copy.
- 1Why anybody wants a clone2 min
- 2The clone process, end to end2 min
- 3Preparing the source without an outage3 min
- 4Configuring the target2 min
- 5What a clone does not copyFree preview2 min
- 6The post-clone checklist3 min
- 7Cloning between different hosts and versions2 min
- 8Refreshing rather than cloning2 min
- 9Making it repeatable2 min
- 10What breaks: four cloning failures2 min
- 11Lab briefing · Clone, then find what you missed2 min
8. Module 7 · Patching
FLAGSHIP. Where most unplanned downtime comes from, and where a rollback either exists or is discovered not to at the worst moment.
- 1Why you are patching at all3 min
- 2Kinds of patch, and what each touches2 min
- 3Reading a readme properly2 min
- 4Online patching: the model2 min
- 5The adop cycle, phase by phaseFree preview2 min
- 6When a phase fails2 min
- 7Abandoning and cleaning up2 min
- 8Rollback: what is actually reversible2 min
- 9Patching the database tier2 min
- 10Testing a patch before production2 min
- 11The outage plan2 min
- 12Patching cadence, and falling behind2 min
- 13What breaks: five patching failures3 min
- 14Lab briefing · Apply, fail, abandon, apply again3 min
9. Module 8 · Concurrent processing
How work actually gets done in EBS, and the single most common thing a user will ring you about.
- 1What a concurrent request is2 min
- 2Managers, queues and workers2 min
- 3Specialisation rules2 min
- 4Work shifts2 min
- 5A request that will never finishFree preview2 min
- 6Reading a request log3 min
- 7Killing a request safely2 min
- 8Managers that will not start2 min
- 9Output, printing and delivery2 min
- 10Purging: the table that ate the database2 min
- 11What breaks: four concurrent processing failures2 min
10. Module 9 · Users, responsibilities and access
Who can do what, why a user cannot see what they expect, and the requests you will be asked to fulfil that you should refuse.
- 1Users, responsibilities and menus2 min
- 2Why a user cannot see a function2 min
- 3Operating units and access2 min
- 4End-dating rather than deleting2 min
- 5The requests you should refuse2 min
- 6Password policy, and what it is worth2 min
- 7Single sign-on, briefly2 min
- 8What breaks: three access failures3 min
- 9Lab briefing · Diagnose four access complaints3 min
11. Module 10 · WebLogic and the technology stack
The layer between the browser and the application. Unfamiliar to most people arriving from the database side, and the source of a whole class of outage.
- 1What sits between the user and EBS2 min
- 2The admin server and the managed servers2 min
- 3Starting, stopping and the order2 min
- 4Reading a WebLogic log2 min
- 5Memory, and the server that keeps restarting2 min
- 6Certificates and HTTPS2 min
- 7Java versions and the technology stack3 min
- 8The load balancer, and who owns it2 min
- 9What breaks: three technology stack failures2 min
- 10Lab briefing · Follow one request all the way down2 min
12. Module 11 · Performance: finding the actual problem
FLAGSHIP. Where a DBA is judged, and where most effort is spent tuning things that were never the problem.
- 1Measure before you tuneFree preview2 min
- 2"The system is slow": turning that into a question2 min
- 3Where time actually goes2 min
- 4AWR and ASH, used properly2 min
- 5A baseline is what makes a report meaningful2 min
- 6Finding the expensive SQL2 min
- 7Execution plans, enough to act3 min
- 8Statistics: the commonest real cause3 min
- 9Locking and blocking3 min
- 10When it is the application, not the database3 min
- 11What breaks: four performance investigations that went wrong3 min
- 12Lab briefing · Diagnose three slow systems3 min
13. Module 12 · Space, growth and purging
Unglamorous, entirely predictable, and the cause of a large share of unplanned outages — because a full disk stops everything at once.
- 1What fills up, and in what order2 min
- 2Monitoring space before it matters3 min
- 3Growing a tablespace2 min
- 4The standard purge programs3 min
- 5Deciding what history to keep3 min
- 6Purging safely the first time3 min
- 7Reclaiming space, and when you cannot3 min
- 8What breaks: three space failures3 min
- 9Lab briefing · Fill it up on purpose2 min
14. Module 13 · Monitoring, and knowing before the client does
The difference between a professional service and a reactive one is who notices first. That is a design decision, not diligence.
- 1What is worth watching3 min
- 2Checking availability honestly3 min
- 3The check that runs on the box it watches3 min
- 4Polling intervals, and what they cannot see3 min
- 5Alerting somebody who will act3 min
- 6Alert fatigue as a design failure3 min
- 7The silent failure: nothing ran3 min
- 8Reporting to somebody non-technical3 min
- 9What breaks: the outage nobody was told about3 min
15. Module 14 · Upgrades and platform moves
A project rather than a task. The DBA work inside it, and the parts where the answer is that this is somebody else's decision.
- 1Upgrade, migration, or move3 min
- 2Why estates fall behind3 min
- 3Assessing what an upgrade will touch3 min
- 4The rehearsal, and how many you need3 min
- 5Timing the outage3 min
- 6The go or no-go decision3 min
- 7Moving to a different platform3 min
- 8Moving EBS to a cloud VM3 min
- 9What breaks: three upgrade failures3 min
- 10Lab briefing · Plan an upgrade you will not run3 min
16. Module 15 · Security, credentials and audit
FLAGSHIP. A DBA holds the credentials to a system containing payroll and bank details. That concentration is the security story, not any single setting.
- 1What you are actually holding3 min
- 2The default accounts3 min
- 3The credential in the context fileFree preview3 min
- 4Changing the passwords that matter3 min
- 5Least privilege for a DBA team3 min
- 6Auditing: what to record and what it costs3 min
- 7Encryption, in transit and at rest3 min
- 8Network exposure2 min
- 9What breaks: four security findings2 min
- 10Lab briefing · Audit an estate you did not build3 min
17. Module 16 · Environments: what test must not have
How many environments, what each is for, and the rule that a test system holding real personal data is a production system with weaker controls.
- 1How many environments, and why3 min
- 2What test must not haveFree preview3 min
- 3Masking, and its limits3 min
- 4Refresh cadence, and who decides3 min
- 5Keeping environments comparable3 min
- 6Access, per environment3 min
- 7Documenting an estate2 min
- 8What breaks: three environment failures2 min
- 9Lab briefing · Write the handover pack3 min
18. Module 17 · When it is not the database
The database is blamed first because it is the part with a name. Establishing where a problem actually is, and saying so without an argument.
- 1Why the database is always blamed first3 min
- 2Ruling the database in or out quickly2 min
- 3When it is the network3 min
- 4When it is storage3 min
- 5When it is the client3 min
- 6When it is a person3 min
- 7Writing a finding somebody will accept3 min
- 8What breaks: two arguments that went badly3 min
19. Module 18 · Capstone: a system handed over
One estate taken from inherited and undocumented to something somebody else could run, with every module in this course applied where it belongs.
- 1The brief2 min
- 2Establishing what you have2 min
- 3Making it recoverable3 min
- 4Making it observable3 min
- 5The performance complaint2 min
- 6The handover3 min