Skip to content

Instantly share code, notes, and snippets.

@lukaszstolarczuk

lukaszstolarczuk/LibpmemSeqR.fio

Last active January 20, 2022 04:40
Show Gist options
  • Star 2 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save lukaszstolarczuk/b358293ad818447f0f0388161bbaa332 to your computer and use it in GitHub Desktop.
Save lukaszstolarczuk/b358293ad818447f0f0388161bbaa332 to your computer and use it in GitHub Desktop.
Download ZIP
Example of FIO workload for libpmem engine
Raw
LibpmemSeqR.fio
#
# run command for this workload:
# numactl -N 0 fio LibpmemSeqR.fio # it binds this FIO workload to CPU 0 only. It should be set according to /dev/pmem6 socket's number
#
[global]
thread
group_reporting
norandommap
overwrite=1
thinktime=0
sync=0
direct=0
ioengine=libpmem
iodepth=1
#
# Preparing namespace with commands:
# mkfs.ext4 /dev/pmem6
# mount /dev/pmem6 /mnt/pmem6/ -o dax
#
directory=/mnt/pmem6 # I'm working on directory; each of 'numjobs' gets its own file to work with
size=50GB # Each file will have this size; so in summary they will fill the whole namespace
unlink=0
time_based
runtime=300
ramp_time=30
[SeqRead]
bs=1M
numjobs=16 # depending on CPU core count more can be better, but also can be a bottleneck
rw=randrw
rwmixread=100
percentage_random=0
@jrgruher
Copy link

Does the libpmem io engine support io depths other than 1?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment