Hello Everyone, I am working on determining wear leveling behavior on SSD. The request to me is to run load for a set period of time, we can use 24 hrs as an example, and at the end of the run compare the performance/latency for each hour. So in theory any graph would have up to 24 sets of entries on it. My initial though was to run fio and then in a different window run a script that just copied the logs off every 60 minutes. Im using the below for the configuration file (this was a 2 hour proof of theory run not the full length run) and Im just not seeing any log files to copy. [global] bs=512k direct=1 ioengine=libaio iodepth=16 filename=/dev/sdc ; or use a full disk, for example /dev/sda runtime=120m time_based [seqwrite-phase] stonewall rw=write fill_device=1 write_bw_log=sdc-iodepth16-seqwrite-bs512k-1 write_lat_log=sdc-iodepth16-seqwrite-bs512k-1 write_iops_log=sdc-iodepth16-seqwrite-bs512k-1 For the current version of fio is there a way to gather the logs every X# of minutes? IE a command that tells fio to keep the logs flushed out or to dump the logs every X# of minutes? For the feature request would it make sense to have the ability to either increment the log file every X minutes or to put a time stamp of some sort in the log file so you can split it on that stamp? Thanks, Roger -- 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
