Hi all, I have made an I/O trace which includes "read", "write" and "wait" " commands. When I replay the trace with high I/O depths (for example 32) I see around 10s (!!!!) of latency for completion of 99% of reads and 6.5s latency for the writes. Does anyone know what exactly it means and what is wrong here? It is also notable that the average wait time between I/Os is around 14ms and its median is 8ms. Here is my job file: [global] ioengine=libaio direct=1 iodepth=32 log_offset=1 read_iolog=temp write_bw_log=replay write_lat_log=replay write_iops_log=replay [job] and here is a part of my trace: fio version 2 iolog /dev/sdb add /dev/sdb open /dev/sdb wait 0 512 /dev/sdb read 52321808896 36352 /dev/sdb wait 89 512 /dev/sdb read 52478476288 131072 /dev/sdb wait 192 512 /dev/sdb read 52499185664 131072 /dev/sdb wait 208 512 Thanks in advance -- To unsubscribe from this list: send the line "unsubscribe fio" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
