Download/Upload very slow

I have dedicated server with LXD containers. Seafile servers (ver. 7 and 9) are in separate containers.
Disks are HDD in RAID5.
Mother server still has free resources I mean 5 GB free RAM and unused CPU.
There aren’t any other traffic on this mother server when I’m testing any Seafile server and its bad. I’m just trying to download ISO file about 300MB and transfers are as following:

And the problem is that speed of these Seafile servers is very poor (about 500KB/s download and the same and very unstable upload), on both 7 and 9 version. What can cause that situation ?

And these poor transfers are bad both on Seadrive client and in webbrowser.

And its worth to mention that in one of these containers I have FTP server which has brilliant speeds (maximum of this dedicated server so about 1Gb/s and 250Mb/s upload).

I’ll be appreciate for help.

I’ve found the reason heh.

That was network tuning which I implemented maybe 2 months ago (in sysctl.conf) on mother server:

# Increase number of incoming connections
net.core.somaxconn = 4096
# Increase number of incoming connections backlog
net.core.netdev_max_backlog = 65536
# Increase the maximum amount of option memory buffers
net.core.optmem_max = 25165824
# Increase the tcp-time-wait buckets pool to prevent simple DOS attacks
net.ipv4.tcp_max_tw_buckets = 1440000
net.ipv4.tcp_tw_reuse = 1
#Number of times SYNACKs for passive TCP connection.
net.ipv4.tcp_synack_retries = 2
#Allowed local port range
net.ipv4.ip_local_port_range = 2000 65535
#Protect Against TCP Time-Wait
net.ipv4.tcp_rfc1337 = 1
#Decrease the time default value for tcp_fin_timeout connection
net.ipv4.tcp_fin_timeout = 15
#Decrease the time default value for connections to keep alive
net.ipv4.tcp_keepalive_time = 300
net.ipv4.tcp_keepalive_probes = 5
net.ipv4.tcp_keepalive_intvl = 15
net.core.rmem_default = 65536
net.core.wmem_default = 65536
net.core.rmem_max = 16777216
net.core.wmem_max = 16777216
net.ipv4.tcp_mem = 50576 64768 98152
net.ipv4.tcp_rmem = 4096 87380 16777216
net.ipv4.tcp_wmem = 4096 65536 167

Well… some of these parameters broken network to LXD containers and thats why speed was poor. Now is excellent.

Cheers all especially Seafile team :slight_smile: