Hi all,

till a few days ago we were noticing regular build failures in Jenkins. I think 
I might have found and fixed the problem we were having.
At the beginning of a new build, we usually clean the workspace. Unfortunately 
this is only done on the primary node. There are two steps where we have to do 
things on other nodes.
The main problem was the deploy step … here we were “unstashing” all the 
artifacts we want to deploy. Unfortunately this was not replacing the content 
of the local directory, it was unpacking the stash.
This resulted in 10-20 versions of each file being deployed on every run. This 
was also why the builds kept on getting longer and longer. I had to search for 
the right place to force a cleanup and I think the current version works … at 
least it’s saving us 10-20 minutes on the deploy phase. … Keep your fingers 
crossed, that things will be staying that way.

Chris

Reply via email to