load snapshot fails on low end machine
Wile loading a snapshot it failed after 3 hours with OOM-killer on a machine with 8GB RAM machine. While it's an edge case and we recommend way more RAM for hived, it's surprisingly low performance on low end machine. I.e. copying shared_memory.bin file from a different node will allow to hived to start there. Possibly related to rocksdb configuration to run so many threads. So maybe --lowmem option?