Hi all, ok … well yeah … I however have noticed that the redeploy functionality of the deployAtEnd still aborted on some errors. So possibly the 503 could be one of these.
In PLC4X we had to use the two-step process as we must deploy from a different machine as we build, but I also recall the “deployAtEnd” having its issues but not quite what they were. Ok yeah … that form of deploying SNAPSHOTS actually is quite bad for Apache’s Nexus … May I ask why you are doing that? Should this sort of be some replacement for a “weekly automated release”? Chris Von: Jan Šťastný <[email protected]> Datum: Mittwoch, 22. Mai 2024 um 10:55 An: [email protected] <[email protected]> Betreff: Re: [HEADS-UP] SNAPSHOTS are not up to date Hello Christopher, we're actually switched from the local deployment recently to -DdeployAtEnd with a failover of 5 times, which should be really similar. Currently we're getting HTTP error 503 which we haven't seen before, it's a proxy error, not timeout. And regarding our SNAPSHOT policy, yeah, we're doing something extra - except from 999-SNAPSHOT which is our main branch snapshot version, we do 999-YYYY-MM-DD-SNAPSHOT snapshot deployments, we call it weekly (but seems to be more often even), which results in many snapshot versions being kept next to the 999-SNAPSHOT one, so at the moment it's as if we had 30 snapshot versions. Example is here: https://repository.apache.org/content/groups/snapshots/org/drools/drools-bom/ - see 999-SNAPSHOT is there, having just like 6 builds inside that folder, but outside of 999-SNAPSHOT folder there are already many directories (other snapshot versions) e.g. 999-20240517-SNAPSHOT - those I am afraid don't fall under existing retention policies for snapshots as they contain only a single build each and are kept there for quite some time. Regards Jan
