One issue with traditional database systems like PostgreSQL is, that you cannot easily scale vertically. Of course you could add read replicas and offload read operations, but that either requires changes in the application, or putting something in front that understands the PostgreSQL dialect and automatically routes writes to the primary and spreads reads across one or more replicas (e.g. pgpool-II). But even if you have something in place, you need to deal with replications…
As a consultant at dbi services we can dedicate a significant part of our time to training and I took the opportunity today to attend an online conference. In this blog post, I will present this event and one of the sessions that particularly interested me. The event I followed is the Virtual Symposium – SQL Server & Azure SQL organized by SQLServerGeeks. It’s a free event that can easily be followed on Zoom but…
There is a lot of roumors around since it was anounced to switch CentOS 8 from a Downstream Build from RHEL 8 to a Upstream Build for RHEL 8 on December 8th 2020. Many people asked why a second Upstream to Fedora and the way of communication choosen was sub optimal in diplomatic words. Cutting the support date from 31st of May 2029 to 31st of December 2021 brings many people in trouble which has…
By Franck Pachot . Delphix is a tool for easy cloning of databases. The idea is that all is automated: the user can create a clone, rewind or refresh it with one click. However, I was suprised that the following common scenario is not managed by the Delphix engine: You clone from production, say Oracle 12c You upgrade the clone, say Oracle 19c You test there You refresh the clone from production, obviously being back…
In this blog I am describing the steps to patch an ODA HA from the version 19.8 to version 19.10. The patching was done with a customer. Thanks to Pierre. The name of the servers were changed The ODA is an X8-HA with following nodes -server-aa -server-ab Before starting the patches, be sure that you have enough space on / /u01 and /opt filesystems. Below the state of our filesystems [[email protected] Downloads]# df -h /…
PostgreSQL comes with a rich set of administrative functions for various use cases. There are functions to control parameters, there are functions for index maintenance, others are for recovery and or backup control and there is much more. In this post we’ll look at how you can use the build-in functions to set and get parameters without touching the configuration file (or setting parameters at the cluster, database and user level). In the posts that…
Testing the latest features or bug fixes in PostgreSQL can easily be done by pulling the latest commits from the git repository and compile PostgreSQL from source code. I am doing that frequently on my local sandbox VM but this becomes limited when I need more CPUs or more memory. This is where the public cloud really shines, as you can easily deploy new VMs, do your tests and get rid of it afterwards with…
Introduction In an Oracle database, services have multiple purposes: identifying group of users prior authenticating them enabling/disabling access to a database for particular groups managing preferred nodes on a RAC configuration redirecting users to the primary after a switchover/failover when using Data Guard or Dbvisit Standby redirecting read only connections to Standby database with Active Guard option A customer asked me if services could be used for switching between PDBs without modifying client’s connexion, for…
By Franck Pachot . Oracle Error: ORA-01547: warning: RECOVER succeeded but OPEN RESETLOGS would get error below ORA-01152: file … was not restored from a sufficiently old backup ORA-01110: data file … This error is one of the most misunderstood recovery error. It can happen in different case but I have here a simple example to reproduce it, with some comments and queries to see the state. I run an example to avoid long theory…
Introduction When performing an audit on an SQL Server VM the first thing the system administrator told me was that there’s no issue with the CPU on this box, the monitoring shows it’s always around 50% usage. Problem Here is the Task Manager: The average CPU usage on this VM is indeed 50% but half of the cores are stuck at 100%. There are two SQL Server instances installed on this server but they are…
Introduction In a previous blog post, I did a T-SQL script reading Audit files to send emails with valuable information about instance changes. In this one, you will see how to get nice emails from critical errors or specific events occurring on the instance using the ErrorLog. A common best practice for any DBA is to configure SQL Server Agent Alert for errors with severity 17 to 25. Alerts are also often set for less…