Running Jetpack components outside of docker

Hi,

I’m currently evaluating Jetpack and was wondering if it’s possible to run XL Release/XL Deploy outside of their containers. If I were to install XL Release standalone (https://docs.xebialabs.com/xl-release/how-to/install-xl-release-as-a-service.html) would a Jetpack license work?

If not, and running in a container is required, would I be able to use an external DB/configuration/plugin folder/etc… with my install?

Thanks,
Brett

1 Like

Hi Brett,

The XL JetPack license is only available when you’re running the components in containers. To run them outside of containers, you’ll need XebiaLabs DevOps Platform licenses, which you can grab here.

Using an external database with the containers is fine; you just have to be sure that the containers can reach the DB. For performance reasons, it’s best to have the containers and the DB in the same network.

You can find info about mounting other directories (like the plugins directory) as volumes here and here.

Thanks, Amy.

I went ahead and mounted the necessary volumes but now I’m running into a strange issue regarding max key length when trying to build the external mysql database.

xl-release_1 | Reason: liquibase.exception.DatabaseException: Specified key was too long; max key length is 767 bytes [Failed SQL: CREATE INDEX IDX_DEPL_STATUS ON xlarchive.DEPLOYMENTS(status)]

I am using standard UTF-8 (NOT UTF8MB4) for all of the tables and mysql aurora 5.6.10a in AWS.

The strangest thing though is when I log into the DB through a normal SQL client, I am able to execute the failed query without generating any error. If I attempt to reinitialize the DB I then get an exception saying the index already exists.

CREATE INDEX IDX_DEPL_STATUS ON xlarchive.DEPLOYMENTS(status)

Have you ever run into the issue before?

Thanks,
Brett

Quick follow up. I was able to sucessfully provision the database with the configuration but using a standard community mysql RDS and not an amazon aurora rds. I’m still unclear as to what is happening, but I can probably make do with this.