PgConf.Russia 2020

MobilityDB is an open source moving object database system (https://github.com/ULB-CoDE-WIT/MobilityDB). Its core function is to efficiently store and query mobility tracks, such as vehicle GPS trajectories. It implements the Moving Features specification from the Open Geospatial Consortium (OGC). MobiltyDB is engineered up from PostgreSQL and PostGIS, providing spatiotemporal data management via SQL. It thus integrates with the postgreSQL eco-system allowing for complex architectures such as mobility stream processing and cloud deployments.

The presentation will explain the architecture of MobilityDB, its database types, indexes, and operations. We will highlight the PostgreSQL features that enable this extension, and the would like to have features. This presentation will be of special interest to the PostgreSQL community, and to professionals in the transportation domain.

У меинтейнеров open-source возникает множество проблем по мере их роста. Как писать все больше требуемых фич, чинить все больше issues'ов и успевать смотреть все больше pull request'ов? На примере WAL-G(backup-tool for PostgreSQL) расскажу про то, как мы решали эти проблемы, запустив курс по Open-source разработке в университете, чего мы добились и куда будем двигаться дальше.

В докладе мы хотим рассказать об опыте миграции одной большой системы с Oracle на PostgreSQL. Сама система построена на стеке PHP + Oracle, её отличительной особенностью было то, что вся бизнес-логика была реализована в PL/SQL коде. В СУБД - это более 3000 пакетов с 4-10 функциями (процедурами) в каждом. В PHP - более 10000 форм со вставками анонимных блоков, используемых для получения данных, обработки и сохранения результатов в Oracle.

Для решения этой крайне объемной работы мы взяли ANTLR4 (синтаксический анализатор для чтения, обработки, выполнения или перевода структурированного текста), грамматику PL/SQL и создали автоматический конвертер, который позволяет преобразовать все объекты схемы и нашей системы из Oracle в рабочий код для PostgreSQL.

Cloud storage has some unique characteristics compared to traditional storage mainly because it is virtualized and controlled by software. One example is that AWS EBS shows higher throughput with larger I/O size up to 256 KiB without hurting latency. Hence, a user can get only about 4 MiB/sec with 1,000 IOPS EBS volume if the I/O request size is 4 KiB, whereas a user can get about 250 MiB/sec if the I/O request size is 256 KiB. This is because EBS consumes one I/O in a given IOPS budget for every I/O request regardless of the I/O size (up to 256 KiB). Unfortunately, PostgreSQL cannot exploit the full potential of cloud storage because PostgreSQL has designed without considering the unique characteristics of cloud storage.

In this talk, I will introduce the AppOS extension that improves the throughput of a write-intensive workload by 10x by transparently making PostgreSQL cloud storage-native. AppOS works like a storage driver that efficiently exploits the characteristics of cloud storage, such as I/O size dependency to storage throughput and latency, atomic write support in cloud block storage, and fast, but non-durable local SSDs. To do this, AppOS comprises a Linux-compatible file I/O stack including virtual file system, page cache, block I/O layer, cloud storage driver. On top of the file I/O stack, syscall module supports registering pre- and post-handler for file I/O-related system calls in order to transparently work without modifying PostgreSQL codes.

I will focus on presenting key use cases and performance results of the AppOS extension after explaining the internals. Specifically, I will show the performance results of OLTP and some batch workloads using standard benchmarking tools like pgbench and sysbench. I will also present performance results and implications on multiple clouds including AWS, GCP, and Azure.