PgConf.Russia 2020

I will share our experience of migrating a server application for Russian railways from Oracle 11g Standard Edition to vanilla PostgreSQL 11.5.
At the time of migration, the database contained about 200 stored procedures of over 60000 lines of Oracle PL/SQL code (which has been developed since 2006, that is, for more than 12 years), about 250 tables, and 50GB of data.
Starting with a prologue, we'll describe our adventures along the migration process, as well as pleasant and unpleasant surprises we encountered, and finally get to an epilogue and a happy end. The story is told on behalf of an Oracle user exploring Postgres.

When migrating data from one DBMS to another, the question arises: choose a third-party tool or to program the migration yourself? Companies, trying to grow competencies within themselves, choose the second option. And they come across the "invention of their own bicycles". However, the market has powerful free data migration tools. One such tool is Pentaho Data Integration, part of the Pentaho Community Edition. The report will discuss the use of this package for data migration between Oracle and PostgreSQL. Particular attention will be paid to the problems with using this tool, and to the tasks of testing for the completeness and integrity of migrated data.

Small video illustration:

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.

At the master class, we will learn how to execute fast load tests of Postgres databases: optimizing database configurations, data structure, indexes, OS settings, etc. To do this, we will create a code, build the infrastructure for the test from it and will do the test. As a result, we get a flexible tool in the code to which you can attach any monitoring, and for which you don't have to pay a lot of money, because the environment is created in 7 minutes in an empty AWS account and destroyed after test