PostgreSQL Conference Asia Pacific - PGConf APAC 2018
III международная конференция пользователей и разработчиков СУБД PostgreSQL стран Азиатско-Тихоокеанского региона состоится в Сингапуре с 22 по 23 марта 2018 года.
Конпания Postgres Professional участвует в этой конференции и будет представлена тремя докладчиками:
► Олег Бартунов и Александр Коротков выступят с докладом «In Postgres we Trust» и расскажут об интеграции СУБД с технологией блокчейн, а так же представят новый проект компании - «Credereum» - созданный специально для поддержки такой интеграции.
In Postgres we Trust - Oleg Bartunov
Bringing the provability and immutability of blockchain to performance and efficency of traditional DBMS.
Blockchain technology has several unique properties including provability and immutability. Every blockchain transaction is signed by its author, and it could be verified by any blockchain network member. Also, once data is stored in blockchain, it can’t be altered in the future. Many databases operating traditional DBMS would also benefit from provability and immutability properties. However, inclusion of all the transaction data in the public blockchain is very expensive.
Credereum is the platform, which allows creation and maintaining of databases, whose contents and history are provable and immutable without sacrifice the performance and efficiency of traditional DBMS. Thanks to Credereum, database owner can prove the validity of query results, while users can verify them. Database owner don’t have to reveal the whole database contents or full history of transactions to provide the proof of database query results. Therefore, Credereum database may contain private sensitive information. Credereum utilized bleeding-edge technologies including, but not limited to decentralized cloud, public blockchain with sharding. Credereum is an emerging technology of trusted and private databases.
We will explain why PostgreSQL is suitable database for Credereum and what we need to develop in Postgres to support signed transactions and cryptographic storage.
Анонс доклада в программе конференции
► Разработчик-исследователь компании Антон Быков выступит с докладом «Realtime no-sql data processing and analytics in PostgreSQL».
Realtime no-sql data processing and analytics in postgresql - Anthony Bykov
It is a usual approach in data science to dump the database and build models upon outdated information. In some cases, it may be unacceptable. So in this session, it will be discussed how PostgreSQL can help on an example of the real project.
First of all, we will discuss the problem data scientists face in their jobs when they want to use PostgreSQL. This problem is unstructured data, which should be somehow stored and accessed. In relational databases, it might be a challenging task, but according to the latest SQL standard, there is a solution. PostgreSQL already provides a special type “jsonb” which can solve the problem.
Tthe next theme of the session is “Real-time data processing”. Nowadays PostgreSQL can handle lots of different connections and data changes. In order to work with actual data, it is necessary to react instantly to any changes. In this part, we will use stored functions and features like “transform” and “plpython”.
Stored functions are really strong tools, but sometimes it is hard to implement the needed feature. So, our next block of the session will be dedicated to plpython and data science models usage on an example of plpython+scikit-learn.
Анонс доклада в программе конференции
Сайт конференции: http://2018.pgconfapac.org