Lucee forcing v2.0.1.25 of the s3 extension

We are trying to use the latest version of the s3 extension. We have historically installed extensions by having a LUCEE_EXTENSIONS env variable. We specify the extension ID, name, and version. We’ve never had a problem until we started trying to install the latest version of the s3 extension (2.0.2.19). We have specified the following…

LUCEE_EXTENSIONS=“CED6227E-0F49-6367-A68D21AACA6B07E8;name=admin;version=1.0.0.5, D4EDFDBD-A9A3-E9AF-597322D767E0C949;name=oracle;version=21.3.0.0000, 66E312DD-D083-27C0-64189D16753FD6F0;name=pdf;version=1.2.0.10, 671B01B8-B3B3-42B9-AC055A356BED5281;name=postgresql;version=42.2.20, 60772C12-F179-D555-8E2CD2B4F7428718;name=redis;version=3.0.0.51, 17AB52DE-B300-A94B-E058BD978511E39E;name=s3;version=2.0.2.19”

I can see as the container starts up that it installs 2.0.2.19 but then “updates” to 2.0.1.25, which doesn’t have a ton of the s3 functions. I can go into the Admin after and update to 2.0.2.19 without a problem, and it even prompts me to update to the latest version, but that doesn’t work very well in a containerized environment.

After digging in some more, it looks like it always installs the 2.0.1.25 version of the s3 extension, even when s3 isn’t listed in our env variable at all. It seems like it is always installed by default by Lucee.


I’m sure I’m doing something silly. How do I stop it from forcing 2.0.1.25 and instead install version 2.0.2.19?

Note - Using the default Lucee images, specifically Lucee:5.4.5.23-nginx

Posted in the cfml slack here - Slack

It looks like this is a bug that was fixed in 6.1. That’s great, but also not great since we are on 5.4. :slightly_frowning_face: Thanks for digging into it on the ticket, Zac.

https://luceeserver.atlassian.net/browse/LDEV-4996
https://luceeserver.atlassian.net/browse/LDEV-5004