First blog entry from the planned series describing how fun it was to prepare my last POUG Workshop 2020 session.
Interesting example of using Jupiter Notebook to visualize performance data gathered from Solaris operating system. In this particular case (real scenario) it allowed us to confirm the issue we were suspecting was related to Solaris scheduler interfering with processor pools and Solaris zones.
I find Jupyter Notebook useful to process, analyze and visualize performance data from operating system, database or any other source. Hopefully my time allows showing some examples of it soon. For now I’m describing how to prepare environment for such analysis – and it involves docker again.
My next post about YUM on Exadata is about creation of local ULN mirror containing Exadata baseline repositories together with generic Oracle Linux channel.
I cover a case when there is no Oracle Linux machine in the environment – so such mirror needs to be created on non-Oracle machine. Specifically I show possibility to use docker image for this purpose.
In the second post I install and upgrade some packages using Oracle public YUM repository.
As a result my database node is no longer consistent with baseline of my firmware.
Let’s find out what will happen when I attempt to install new firmware on it.
First post of the series about usage of YUM on Exadata.
Does Exadata run standard Oracle Linux and therefore one can use YUM generic repositories?
How to do this safely without breaking dependencies or risking problems during next patching window.
Why, the heck – another technical blog?