I have a MongoDB deployment, where all databases are around 500GBs in size (replicaset / no sharding). Testing the PBM, creating a full backup, it took around 17h to create a compressed backup (GZIP), which is a little bit better than “mongodump”. I was expecting to see some improvements between PBM and mongodump.
One thing I noticed was the fact that the “pbm-agent” process was only using a single core of the 12 available, and I would like to see if there is another way to make PBM use the max number of resources available to perform this backup. Also, is there another parameters that can be setup to tune the overall backup performance?