Identify Acronyms

[sgt-spike]

Can you identify all the acronyms in the read-me?

[oraclesean]

Thanks for your input! I added a glossary to the README—let me know if I overlooked something you wanted to see there.

[sgt-spike]

Thanks for the quick response. After reading through your readme, I went and found the terms and a tutorial about the CBD/PDB. Great stuff.
Do you have a docker pull command for this image?

[oraclesean]

@sgt-spike the images built with this repo are generated locally on the system so there's no need for a pull. You would pull an image from a repository (like Docker Hub) that's already built. Oracle is very defensive about licensing and they're the only ones (legally) allowed to distribute pre-built images that way (unlike pretty much every other DB vendor). :(

With this repo you can build those images and run Oracle databases. At a high level:

• git clone this repo.
• cd (change directory) to the newly downloaded repository directory (eg cd docker-oracle).
• download the database installation file(s) for the version you want to build from Oracle.
• move or copy them into the docker-oracle/database directory.
• make the shell scripts executable on your system: chmod u+x *.sh (chmod changes permissions; u+x says "make the file executable by the user; *.sh is a wildcard meaning "all files that end in .sh, or shell scripts).
• run the build script (I just added this yesterday and haven't documented it yet): ./buildDBImage.sh [version]

When it's complete you'll have two new images—a modified Oracle Enterprise Linux (OEL) with the database installation prerequisites, and a database:

REPOSITORY                                          TAG                  IMAGE ID       CREATED          SIZE
oraclesean/db                                       21.5-EE              9f5b491e022f   29 minutes ago   8.69GB   <<< Database image
oraclelinux                                         7-slim-21c           1f458ed3bc21   40 minutes ago   681MB    <<< OEL + prerequisites

It builds two images because the prerequisites are the same for every minor version (eg 19.3, 19.14), and it takes a long time. I build a LOT of database images and grew tired of sitting through the duplicated effort 3-4 times a day. Having the prerequisite-ready starting point cuts down the time it takes to install/patch the database pieces.

Note that building a "patched" version (like 19.14) means getting the install files to build 19c, then downloading patches for 19.14 from Oracle's Metalink support site (account required). That's not necessary if all you want is a basic database. Look in the manifest files for a given version to see the minimum requirement—the only file you must have is the zip file(s) identified as database in the third column. Everything else shown there is optional (and typically only available through Metalink)—you only need OPatch and patches if you intend to patch the image.

From there, you can docker run the image to start a database container. The first time it runs it needs to create the database, so it will take several minutes. After that it will stop/start in a few seconds.

I'm striving to make this more user-friendly and accessible, but sometimes that's tough. I started working with Oracle "before the turn of the century" (lol) and Docker for 7-8 years. I'm not immune from overlooking what's "obvious" to me, so your input and observations are a welcome and valuable perspective!

[sgt-spike]

I really appreciate your patience and explaining the how to. I do have limited knowledge in docker and have been reading and watching many articles and videos trying to understand from a starting point. You laid out exactly what I was looking for.