No way, the process was eating all memory available to backup a powered off VM which only has 100MB of disk.
I was also getting these warnings in mcserver.log:
WARNING: com.avamar.mc.sdk.McsFaultMsgException: E22234: Group does not exist.
WARNING: com.avamar.mc.sdk.McsFaultMsgException: E22289: Retention policy does not exist.
Finally, it looks like it fixed itself. I only edited a few jobs and it worked again (not sure if related).
Later I also removed all backup jobs, upgraded to 6.1.81, and created jobs again.
Still getting those warnings though.