Oracle FGA internals (Part 4).Complex (column) expression.

On a previous post we considered a case of FGA set for some quite simple condition. Let's set up a new scene for further experimentation/ideation. Clean up old policies and set a new one (that includes condition on the table column/s) begin dbms_fga.drop_policy(object_schema =>'FGA_OWNER' , object_name => 'TBL_FGA', policy_name =>'TBL_FGA' ); end; / begin dbms_fga.add_policy(object_schema …

Oracle FGA internals (Part 3). More complex but still simple condition.

On my previous post we stopped at comparison of noaudit case with a very basic FGA audit setting (actually kind of "No OPeration"). Today I am going to consider slightly more complex case. Let's clean up our audit 1st begin dbms_fga.drop_policy(object_schema =>'FGA_OWNER' , object_name => 'TBL_FGA', policy_name =>'TBL_FGA' ); end; / and set up a …

Oracle FGA internals (Part 2). Simple audit condition.

On my previous post we have set the scene up and raised a couple of questions. Let's start with our new agenda, I have connected to my Oracle session via GDB (debugger). Like gdb -p 12345 and set up a few break points (inside Oracle's codebase) b opifch b selexe0 b expeal Where opifch - …

Oracle FGA internals (Part 1)

Oracle FGA stands for Fine Grained Audit, see for instance here: Official docs or here: Tim Hall's oracle-base site The main question (or concern) for my use case was how switching on FGA on some object/objects may affect executions of queries from an application itselfs that should not be audited. Let's start setting up the …

Terminator to Vagrant box connection

I have the following config piece in my Vagrant file: config.ssh.host = '192.168.1.201' config.vm.network "forwarded_port", id: "ssh", host_ip: "192.168.1.201", guest: 22, host: 25 config.ssh.host port to be used by internal Vagrant ssh config.vm.network "forwarded_port" - VM's port forwarding settings. to connect to this Vagrant box from popular Terminator ssh client I need to point out …

Journey to GPC. Part 2.

My next step was to try to install some open source software and I needed to install git SCM at first. Let's try to do that. My new terraform file looks like: provider "google" { project = "postgretrial" region = "us-central1" zone = "us-central1-c" } data "google_compute_default_service_account" "default" { } data "google_compute_image" "image-terra-io" { provider …

Journey to GPC by means of terraform

By this post I am starting to describe my newbye experience in Google Public Cloud journey. I am going to use terraform project (pre-installed on your Google Cloud shell by the way) as IaC solution. I have collected a set of pitfalls and met many fallacies which I want to share with you. I'll try …

Crunchydata PostgreSQL deployment on GKE step-by-step 2

Once you have crunchydata postgres-operator installed (as described in https://dmitryremizov.wordpress.com/2020/08/20/crunchydata-postgresql-deployment-on-gke-step-by-step/) it is time to configure a new database cluster.First of all you need to setup a pgo client. let's open your "cloud shell" for your new cluster (cluster-3 in my case). It is easy to do from Kubernetes Engine/Clusters page as show below In your …

Crunchydata PostgreSQL deployment on GKE step-by-step

This post is all about  https://www.crunchydata.com/ quite interesting open source project providing PostgreSQL database as managed service. Crunchy data offering exists on GCP (Google Public Cloud) and here I will describe how to set it up step by step from 1st time user perspective. First of all Crunchydata is a part(or partner) of Google Marketplace …

GDB Python API

I've recently realized that gdb have built-in python interface/API see here Debugging with GDB:Python API which can be extremely usefull in investigating Oracle internals. To use python API you need to print "python" keyword in GDB command prompt then any Python code of arbitrary complexcity, finally you should finish you Python subprogram with an "end" …