wal与db分区磁盘util较高

在使用cephfs上层业务进行数据写入的时候，wal与db所在分区的util值较高，可能是由于bluestone的compaction导致，具体排查如下

iotop 查看

查看是否有较多的rockdb：bg0，是的话，代表在compaction

如何减少compaction：

在数据写入之前手动进行触发

ceph daemon osd.0 compact

或者关闭compaction

bluestore_rocksdb_options = num_levels=4,compression=kNoCompression,max_write_buffer_number=4,min_write_buffer_number_to_merge=2,recycle_log_file_num=4,write_buffer_size=33554432,writable_file_max_buffer_size=0,compaction_readahead_size=65536,level0_file_num_compaction_trigger=16,level0_slowdown_writes_trigger=128,level0_stop_writes_trigger=128,bytes_per_sync=4194304,wal_bytes_per_sync=1048576,optimize_filters_for_hits=1,bloom_locality=1,skip_stats_update_on_db_open=true,hik_compaction_delay_ms=600,max_subcompactions=1,disable_auto_compactions=true