jezik/language:language

Letos bo DBA področje namenjeno različnim tematikam. Govora bo tako o klasičnih DBA vsebinah kot tudi o novejših pristopih za časovno učinkovito upravljanje in vzdrževanje podatkovnih baz in povezanih okolij.

Predavanja bodo pokrivala tematike varnostnega kopiranja podatkov, analize in izboljševanja performans podatkovne baze in aplikacij, dolgotrajnega hranjenja in analize preteklih in trenutnih performančnih lastnosti podatkovnih baz, novosti v verziji 12.2, avtomatizacije ponavljajočih nalog DBA-jev in sistemskih administratorjev, predavanja glede nadgradenj na novejše različice Oracle podatkovne baze, obravnavali bomo GDPR regulativo za poenotenje in zaščito podatkov posameznikov na območju EU in izvoz podatkov izven EU območja in še kaj se bo našlo.

S predavanji bomo tako pokrili širok spekter tematik. Med letošnjimi novostimi je dan namenjen predavanjem in delavnicam. Zadnji dan srečanja, t.j. v sredo, bomo tako poleg predavanj imeli tudi delavnico na kateri bo prikazano kako lahko z orodjem Delphix v trenutku, torej neodvisno od velikosti podatkov, izdelamo popolno delujoče klone želene podatkovne baze iz stanja v izbranem času. V tej delavnici bo govora tudi o pomembnosti zagotovitve testnih podatkov ustreznih velikosti in vrednosti, pa tudi kako lahko te podatke po potrebi maskiramo, če je osnova produkcijska podatkovna baza.

 

Vrstni red predavanj je urejen po alfabetičnem vrstnem redu naslovov predavanj.

 

Arhitektura Oracle in SQL Server podatkovnih zbirk iz vidika restavriranja podatkov

Urh Srenik
engUrh Srečnik
Abakus plus d.o.o.

 

Abstract

V Abakusu kot Oracle DBA-ji večkrat naletimo na situacijo, da ima stranka več kot en tip podatkovne zbirke. V SQL svetu se kot konkurenca Oraclu najpogosteje pojavlja Microsoft SQL Server. Ker smo z našim produktom Deja Vu želeli zadostiti zahtevam strank, da bi na enoten način izvajali varnostno kopiranje različnih podatkovnih zbirk, smo letos za ta produkt razvili polno podporo tudi za SQL Server.

V letu dni, odkar smo začeli s tem projektom, smo se precej podrobno seznanili z razlikami in izzivi, ki jih različna tipa podatkovnih zbirk ponujata. Na predavanju bomo iz arhitekturnega in praktičnega vidika predstavil, kakšne te razlike so in kakšne prednosti prinaša eden ali drug.

Namen predavanja ni iskati boljše/slabše implementacije rdbms-ja. Namen je, da predvsem Oracle DBA-ji, ki imajo manj izkušenj z drugimi podatkovnimi zbirkami, dobijo vpogled v drugačen, še ne raziskan svet. Poleg tega bodo izkušenejši DBA-ji nedvomno dobili precej idej/vprašanj glede posledic, ki jih različne implementacije prinašajo. Te DBA-je po predavanju vabimo na pivo in nadaljnjo diskusijo.

Automate DBA Tasks With Ansible

Ivica Arsov

 

Abstract

Every time you find yourself doing the same thing over and over, have you asked your self: How can I automate this? Nowadays cloud vendors are a major driving force for automation in our industry.

This presentation focuses on the increasing role of automation and how it impacts the way Database Consultants and Administrator job is evolving. You will learn how you can decrease boring tasks and free up time for value-add work. We will take a detailed look at Ansible framework with which we can automate repetitive tasks.

AWR Warehouse: the Good, the Bad and the Ugly

Patrick Jolliffe
engPatrick Jolliffe
Li & Fung Trading Ltd

 

Abstract

In the presentation we will start with some a high level overview of the functionality provided and how it enhances the feature-set of Enterprise Manager. We will discuss the licensing advantages of offloading historical AWR data and processing into a separate database instance.

Then we will review how the ETL process is actually implemented, particularly the tables which store metadata for the extract and load portions. Build on that we will provide some ideas about how to troubleshoot when things don't work as expected, and how to workaround some of the limitations of the product.

Back to the basics: TABLESPACES… but a little bit low level style

Kamil Stawiarski
engKamil Stawiarski
ORA-600 | Database Whisperersace director

 

Abstract

There are a lot of myths and legends about storing segments in tablespaces - should we keep indexes separated from the tables? Can we stop worrying about segment fragmentation? What is real difference between MOVE and SHRINK?

At this session I will present you my free C++ tool to visualize the contents of the datafile. We will disassemble a database block and try to understand how Oracle is really managing space inside tablespaces.

Common Pitfalls in Complex Apps Performance Troubleshooting

 

Abstract

Faced with a performance issue in complex systems people tend to look at things they find easy to get and interpret: CPU & IO usage, Load, Wait Events. The numbers may help in finding what is going wrong - *sometimes*, and hide the real issue easily.

This session is going to discuss simple things to remember while tackling performance issues in complex applications deployed in multiple tiers.

Deadlocks

Peter ampa
engPeter Čampa
Četrta pot d.o.o.

 

Abstract

Predavanje bo na temo deadlock-ov. Pogledali si bomo, v katerih situacijah pride do deadlock-ov, kaj nam sporoča tracefile, ki ga generira deadlock in kako ga pravilno interpretirati. Na praktičnih primerih bomo prikazali, kako pridemo do deadlock-ov in kaj lahko naredimo, da do njih ne pride več.

Dino DBA goes Dev-Ops

 

Abstract

As a DBA or Database-person you will be asked to join a Dev_Ops team and use some of the Agile methods. When this happens, you will have to absorb new knowledge and work-habits. But you will also have to educate your colleagues on what is possible and impossible at the database level. We will try to give some guidelines, and point out a few embarrassing mistakes we made in these situations.

dNFS for DBA's

 

Abstract

Direct NFS is an Oracle implementation of the standard NFS protocol inside Oracle Database. Using this native solution Oracle instance is able to better handle IO traffic and bypass a kernel layer of operating system. dNFS is commonly used for Exadata backups into ZFS Appliance and for Oracle databases using a NFS based solutions (like NetApp or Delphix ).

The goal of this presentation is a show DBA’s what is a Direct NFS, how to configure, use and monitor it. Proper configuration and monitoring are important to achieve a maximum performance and easy problem diagnostic. There will be examples of configuration with and without direct NFS.

NFS protocol is based on the TCP/IP protocol and this presentation will also show a recommended configuration of the TCP stack and explain how network should be monitored.

Ensure Performance Stability when Upgrading Databases

Mike Dietrich
engMike Dietrich
Oracle Corporation

 

Abstract

The biggest fear of DBAs upgrading or migration database is bad performance afterwards.

This presentation will guide you through a simple and straight forward approach to ensure performance stability when upgrading or migrating Oracle database from one release to another. It will also talk about know pitfalls and how to avoid them successfully.

Implementing incremental forever strategy for Oracle database backups

Ilmar Kerm
engIlmar Kerm
Kindred Groupace director

 

Abstract

When databases get ever larger and larger, backing them up using traditional RMAN backupsets will quickly get unfeasible. Completing a backup requires too much time and resources, but more importantly the same also applies to restores. RMAN has always provided a solution as incrementally updated image copies, but they are much less manageable than backupsets.

This presentation goes into detail on how to successfully implement incremental-forever database backup strategy together with a capable storage system like Netapp or ZFS Storage Appliance, and on how to build more advanced features on top that platform, like automation and cloning.

It's a Disaster! Making it better with FarSync and Fast Connect Failover

Neil Chandler
engNeil Chandler
Chandler Systemsace director

 

Abstract

When sat in a meeting discussing Recovery Point Objectives and Recovery Time Objectives, the CIO will always say "No Downtime! No Data loss!"

What is the most effective way of approaching this hard-to-achieve goal? How did we minimise downtime and data loss?

This presentation shows how we reached the decision points for the application, which options were discarded and the final implementation of technology to approach the requirements.

And what simply was not possible without application re-engineering.

Kakšno varnostno kopiranje potrebuje vaša podatkovna baza?

Primo Bedra
engPrimož Bedrač
Oracle Software d.o.o.

 

Abstract

Pristopov k varnostnemu kopiranju je Oracle podatkovne baze je veliko, kateri izmed njih je najbolj ustrezen pa vedno predstavlja svojevrsten izziv. Oracle kot proizvajalec podatkovne baze in strojne opreme ponuja tudi rešitve za njeno varnostno kopiranje. Kot proizvajalec podatkovne baze je Oracle v določeni prednosti, saj lahko ponudi tudi koncepte in tehnologije, ki jih ostali proizvajalci ne morejo. V zadnjem času se je obstoječim rešitvam pridružila še možnost varnostnega kopiranja v Oracle Cloud.

Little Red Riding Hood and GDPR : Who’s afraid of the big GDPR

Philippe Fierens

 

Abstract

Wikipedia states :

The General Data Protection Regulation (GDPR) (Regulation (EU) 2016/679) is a regulation by which the European Parliament, the Council of the European Union and the European Commission intend to strengthen and unify data protection for all individuals within the European Union (EU). It also addresses the export of personal data outside the EU. The primary objectives of the GDPR are to give citizens and residents back control of their personal data and to simplify the regulatory environment for international business by unifying the regulation within the EU.

What does this mean for the Oracle DBA how can he help his company to be compliant with this regulation ? Which Oracle products can help you with this difficult task. This session will be a mix between practical experience and theory. Both speakers worked at a highly secured complex banking environment

Oracle 12.2 Datatabase - novosti

Robert Koroec
engRobert Korošec
Oracle Software d.o.o.

 

Abstract

Na verzijo Oracle 12.2 smo v Oracle ekosistemu čakali nekaj let in sedaj je čas, da si podrobneje pogledamo, katere novosti prinaša Oracle 12.2 primarno za Administratorje kot tudi tudi za razvijalce za čim boljšo izrabo vseh funkcionalnosti v Oracle 12.2

Revizijska sled z Flashback Data Archive

Sunil Jayasinghe
engSunil Jayasinghe
Nova KBM d.d.

 

Abstract

Predavanje o zelo aktualni temi hrambe in sledenju spremembam v bazi podatkov. Ena od obveznosti finančnih institucij (in tudi drugih) je hramba pomembnih podatkov ter zagotavljanje takoimenovane revizijske sledi. Flashback Data Archive (FDA) funkcionalnost baze 11g omogoča samodejno shrambo vseh sprememb na izbranih tabelah za daljše časovno obdobje. Vse to se dogaja v ozadju brez intervecije s strani uporabnikov ali aplikacij. Arhivski podatki so dostopni z navadnimi SQL ukazi. Z uporabo FDA dokumentiramo zapisovanje in spreminjanje zapisov v celotni življenjski dobi revizijske sledi, od izvornega zapisa do trenutno veljavnega zapisa. Iz revizijske sledi je razvidno, kakšna operacija je bila izvedena nad posameznim zapisom, s katerimi podatki in kdaj. Od Oracle verzije 12g pa dodatno še podatek o tem, kdo je spremembo izvedel .

Na predavanju bomo predstavili:

  • opis tradicionalne in najpogostejše metode sledenju spremembam z uporabo triggerjev v bazi,
  • predstavitev napredne metode z uporabo FDA,
  • konfiguracija in vklop FDA,
  • primer uprabe FDA,
  • prednosti/slabosti obeh metod,
  • kompresija zgodovinskih tabel (vrstic, LOB polj, deduplikacija LOB polj),
  • izvoz/uvoz zgodovinskih tabel.

So, my query plan says "Table Access Full" - what happens next?

 

Abstract

For some reports the fastest way to get them done is to use an index to go straight to the rows of interest, but sometimes the only way that makes sense is to scan the entire table. I'm the Oracle Architect for Smart Scan and the engineer in Oracle co-responsible for the table scan driver and this talk will show what happens under the covers in simple easy to understand steps.

We will look at (1) what is a granule in PQ (2) how exactly does the system decided whether to use buffer cache to direct read, (3) how is physical I/O done including Multi-Block reads and Async-I/O (4) what wait events and stats to look at (5) what tuning options are available to the DBA, and (6) for the truly curious how to trace what is happening.

SSD in Oracle

Sergej Roman
engSergej Rožman
Abakus plus d.o.o.

 

Abstract

Približno 10 let se v podjetju Abakus plus že srečujemo s SSD diski, ko smo v enem od poznanih podjetij nanje prvič namestili Oracle zbirko podatkov. O SSD diskih se je od takrat nabralo ogromno izkušenj. Začetni vtisi so bili zelo navdušujoči, sčasoma pa so se pokazale tudi nekatere slabosti.

Največja prednost uporabe SSD diskov je, da zbirkam podatkov zagotovijo neverjeten pospešek. Tudi cena je že zelo sprejemljiva. Slabosti naj zaenkrat ostanejo skrivnost in jih bomo razkrili na predstavitvi. Predstavili bomo tudi nekaj dobrih praks, s katerimi se težavam lahko vnaprej izognemo.

V splošnem pozitivne izkušnje prevladujejo nad slabostmi in danes SSD diske uporabljamo praktično povsod. Nedavno smo s SSD diski v infiniband SAN diskovnem sistemu lastne proizvodnje dosegli nove rekordne rezultate. Paket iz zbirke podatkov dbms_resource_manager.calibrate_io je pokazal pretočno hitrost prek 3 GB/s in 150.000 IO operacij na sekundo pri zakasnitvi 0 ms.

The journey of a BI database stack to the cloud

 

Abstract

BI is and can be a complex matter in big organisations. This presentation will tell the story about a strategic business decision to move the on premise BI stack to the Microsoft Azure cloud. The main focus is on how the Oracle Database is moved as the application servers were moved using a lift and ship mechanism.

This presentation will cover: - how we managed to move +300TB divided over several multi Tb databases to the cloud, - which services were chosen to run the database servers and why, - how performance testing and sizing has been done - how automatic provisioning has been done for providing developers what they need in agreed SLA's, - how we dealed with the short deadline, - why the Azure Cloud was chosen over the Oracle cloud, - Cost savings?

The deadline of the project was end of the year 2016. This gives the opportunity to give you an insight in a very fresh project of the move to the cloud for a complete BI environment.

The new Oracle Release and Update Model

Mike Dietrich
engMike Dietrich
Oracle Corporation

 

Abstract

As of July 2017 Oracle changed its release and update model for the Oracle Database. But not even the release numbering changes, the entire schedule and naming for patches, BPs and PSU changes as well.

This presentation will shed some light on the new release and update model including recommendations for patches and how to find the most important patches for your environment.

Upgrade to Oracle Database 12.2 - Live and Uncensored

Mike Dietrich
engMike Dietrich
Oracle Corporation

 

Abstract

Upgrading a database works fine on Powerpoint. But how about a real upgrade?

This presentation will introduce the news around Database Upgrade to Oracle Database 12.2 including the new preupgrade.jar. And the upgrade will be demonstrated live and interactively, on the command line and graphically with the DBUA.

VOODOO - the black magic of cheep cross-database replication

Kamil Stawiarski
engKamil Stawiarski
ORA-600 | Database Whisperersace director

 

Abstract

The history why we wrote our own replication tool between Oracle and PostgreSQL with a bit of low level archivelog architecture explanation.

  • We shuld buy a bar!
  • Of course we should buy a bar!!! Brilliant why we didn't think about it earlier!?
  • Do you have a few hundred thousand euros?
  • A pint then?
  • A pint then.

We all want to have a bar and drive Porshe... But in most of the cases we can afford only a few pints and drive with VW. Replication engines are expensive and buying something like Golden Gate when you only have SE database and you want to replicate it to PostgreSQL is like buying a single malt to make Cola taste more bearable. That's why we wrote our own replicator tool. A simple one. A cheep one. Nothing fancy. But it works. And we will use this subject as an excuse to show some low level archivelog tricks.

What is new in Multitenant 12.2

 

Abstract

Oracle Multitenant was introduced in Oracle 12.1. and as any product of version 1 in Oracle family was very limited in functionality and suffered for many bugs. In 12.2 Oracle added missing functionality to multitenant thus making this functionality much more interesting than it was in 12.1.

In this session we will look at new features of Oracle Multitenant like provisioning capabilities, cloning PDB, refreshable PDBs, relocating PDBs, PDB local undo, flashback single PDB, different character sets in PDBs, improved resource management, Cross-Tenant Aggregation and more. If you had doubts why you should use Multitenant architecture in 12.1 then this session will probably change your opinion.

Why Has My Plan Changed. Top 7 Plan Stability Pitfalls and How To Avoid Them!

Neil Chandler
engNeil Chandler
Chandler Systemsace director

 

Abstract

When you write SQL, Oracle runs it through the Optimiser to determine the fastest way to access the data. This access path sometimes changes without anyone noticing. Sometimes it's faster. Sometimes it's slower. Dramatically, catastrophically slower. Why doesn't Oracle always get it right? Why does it change?

This presentation will answer the question Why - through the use of live demonstrations, and explain what we can do to manage SQL execution plans within the database and minimise the opportunity for Oracle to get it wrong.