akami
-
Posts
8 -
Joined
-
Last visited
Content Type
Profiles
Forums
Downloads
Store
Gallery
Bug Reports
Documentation
Landing
Posts posted by akami
-
-
via docker:
drakkan/sftpgo:
ftp 64s
sftp 50scrazymax/pure-ftpd:
14s
-
With unraid smb all default settings:
sftp is 7.9x faster than smb
sftp also get lower cpu usage
nvme-cow:
TotalRead = 62 MiB
TotalWrite = 62 MiB
TotalFiles = 2,926 (335)
TotalTime = 01:19
TransRate = 0.81 MB/s
FileRate = 36.6 files/snvme-nocow:
TotalRead = 62 MiB
TotalWrite = 62 MiB
TotalFiles = 2,926 (335)
TotalTime = 01:37
TransRate = 0.67 MB/s
FileRate = 30.1 files/s
hdd*1(xfs)
TotalRead = 62 MiB
TotalWrite = 62 MiB
TotalFiles = 2,926 (335)
TotalTime = 01:34
TransRate = 0.69 MB/s
FileRate = 30.9 files/ssmb
win11->unraid(nvme raid1 cache pool)
nvme-cow: 01:19
nvme-nocow: 01:37
hdd*1(xfs): 01:34win11(pc)->win10(pc-vm): 39s
sftp
win11->unraid(nvme raid1 cache pool)
cow=nocow=hdd*1(xfs)=10s -
Default 'virtio-net' opinions provide awful performance.
'virtio-net' also gets much higher cpu usage
VM -> Host
virtio-net :
[ ID] Interval Transfer Bitrate Retr
[ 5] 0.00-10.00 sec 2.25 GBytes 1.93 Gbits/sec 0 sender
[ 5] 0.00-10.00 sec 2.25 GBytes 1.93 Gbits/sec receiver
virtio :
[ ID] Interval Transfer Bitrate Retr
[ 5] 0.00-10.00 sec 30.7 GBytes 26.4 Gbits/sec 0 sender
[ 5] 0.00-10.00 sec 30.7 GBytes 26.4 Gbits/sec receiver
proxmox seems to just use 'virtio' as the default without causing any stability problem
samba share is very slow on nvme cache pool
in General Support
Posted · Edited by akami
correct data
When Docker mounts a path to /mnt/cache/xxx, there is no shfs process usage.
New folder with 70k files
via docker
sftpgo-ftp rclone--transfers=16 11m14.9s
sftpgo-sftp rclone--transfers=16 11m7.4s
syncthing 10m52s
via unraid build-in sftp
host-sftp 2m6s
Transferring many files via docker is still slow, docker's wirte operation is direct to disk. However, via native sftp, data is cached in ram and written to disk.