Hi there! I am author of this blog post.
Large amount of data, combined with using non optimal memory allocation can result in interesting issues in graph database. Thanks to flamegraph and heaptrack, we successfully resolved issues we had in Memgraph concerning high memory consumption and lower performance on LOAD CSV clause.
