We have a Jenkins deployment currently with default memory settings (which per $JENKINS_URL/monitoring seems to mean 2Gb RAM usage overall), and it is set to make weekly full backups which are ZIPped. The installation has a fairly large history of builds (to compare for historic trends etc.) so the ZIP is around 50G lately, and original FULL directories contain approx

During this (nightly/weekend) archivation, the processor is stressed (understandably) and RAM is over the top for hours - see attached screenshots of $JENKINS_URL/monitoring?part=graph&graph=usedMemory and $JENKINS_URL/monitoring?part=graph&graph=cpu.

If jobs start during this time, they tend to die ingloriously: 

> git rev-list --no-walk cb0a37502c7fea2ef1ba74be1fddad37fe6375eb # timeout=10
Executing git in workdir /var/lib/jenkins/workspace/master-evolution@libs/bc724c0d78c043b9428170279bf13c3f124fba0dfa85ad13b1125577e99e540e
java.lang.OutOfMemoryError: Java heap space
Finished: FAILURE 

Is it possible (in future versions) to configure (ideally automatically) or otherwise change the Zip code's use of at least memory in a way that it is also garbage-collected during the archivation?