I am having an issue that I could use a bit of help with. I am running 384.19 on my AC-RT5300, and I have two 8TB drives connected to it; one is used for backups and as a general share drive, the other is intended as a backup for the first drive; I installed opkg on a small partition on the second drive and installed rsync - although I am not actively running backups between the drives right now, I've only tested it to make sure rsync works.
I have a couple macs that are backing up using Time Machine; this works fine, when one of the macs is doing its thing I can watch top and the usage runs around 3.5% of CPU, so no issues there. I also have several win boxes I am backing up using Veeam to an SMB share. The problem is, whenever one of the win boxes is backing up to the samba share, it causes the CPU load on the router to spike; according to top, the single process is using 40% CPU, and I'm attaching screengrabs of the router UI as well as the status of top when it's running ( as well as when it's not.)
I think, although I'm not 100% sure, that when two win boxes are trying to do their thing at the same time, that it locks up the router - SOMETHING is freezing it such that I have to powercycle it to get things working again.
These two are top and usage with a backup running:
And these two are without.
These are the nmbd and smbd processes that are running:
bsoder@RT-AC5300-B6C0:/tmp/home/root# ps w | grep smb
2101 bsoder 5832 S /usr/sbin/nmbd -D -s /etc/smb.conf
2102 bsoder 5524 S /usr/sbin/nmbd -D -s /etc/smb.conf
2118 bsoder 6016 S /usr/sbin/smbd -D -s /etc/smb.conf
5408 bsoder 7144 S /usr/sbin/smbd -D -s /etc/smb.conf
Here's the global section of /etc/smb.conf; I haven't modified this but just in case there's something to look at here. I can share the other subsections if needed.
bsoder@RT-AC5300-B6C0:/tmp/home/root# cat /etc/smb.conf
[global]
netbios name = RT-AC5300-B6C0
server string = RT-AC5300
workgroup = WORKGROUP
username level = 20
unix charset = UTF8
display charset = UTF8
load printers = no
printing = bsd
printcap name = /dev/null
log file = /var/log/samba.log
log level = 0
max log size = 5
security = USER
guest ok = no
map to guest = Bad User
encrypt passwords = yes
pam password change = no
null passwords = yes
force directory mode = 0777
force create mode = 0777
max connections = 5
socket options = IPTOS_LOWDELAY TCP_NODELAY SO_KEEPALIVE
obey pam restrictions = no
use spnego = yes
client use spnego = no
disable spoolss = yes
host msdfs = no
strict allocate = no
wide links = no
bind interfaces only = yes
interfaces = lo br0 192.168.1.1/255.255.255.0
use sendfile = yes
map archive = no
map hidden = no
map read only = no
map system = no
store dos attributes = no
dos filemode = yes
oplocks = yes
level2 oplocks = yes
kernel oplocks = no
wins support = yes
os level = 255
domain master = yes
local master = yes
preferred master = yes
enable core files = no
deadtime = 30
load printers = no
printable = no
max protocol = SMB2
smb encrypt = disabled
min receivefile size = 16384
passdb backend = smbpasswd
smb passwd file = /etc/samba/smbpasswd
Any suggestions as to how to address this?
I have a couple macs that are backing up using Time Machine; this works fine, when one of the macs is doing its thing I can watch top and the usage runs around 3.5% of CPU, so no issues there. I also have several win boxes I am backing up using Veeam to an SMB share. The problem is, whenever one of the win boxes is backing up to the samba share, it causes the CPU load on the router to spike; according to top, the single process is using 40% CPU, and I'm attaching screengrabs of the router UI as well as the status of top when it's running ( as well as when it's not.)
I think, although I'm not 100% sure, that when two win boxes are trying to do their thing at the same time, that it locks up the router - SOMETHING is freezing it such that I have to powercycle it to get things working again.
These two are top and usage with a backup running:
And these two are without.
These are the nmbd and smbd processes that are running:
bsoder@RT-AC5300-B6C0:/tmp/home/root# ps w | grep smb
2101 bsoder 5832 S /usr/sbin/nmbd -D -s /etc/smb.conf
2102 bsoder 5524 S /usr/sbin/nmbd -D -s /etc/smb.conf
2118 bsoder 6016 S /usr/sbin/smbd -D -s /etc/smb.conf
5408 bsoder 7144 S /usr/sbin/smbd -D -s /etc/smb.conf
Here's the global section of /etc/smb.conf; I haven't modified this but just in case there's something to look at here. I can share the other subsections if needed.
bsoder@RT-AC5300-B6C0:/tmp/home/root# cat /etc/smb.conf
[global]
netbios name = RT-AC5300-B6C0
server string = RT-AC5300
workgroup = WORKGROUP
username level = 20
unix charset = UTF8
display charset = UTF8
load printers = no
printing = bsd
printcap name = /dev/null
log file = /var/log/samba.log
log level = 0
max log size = 5
security = USER
guest ok = no
map to guest = Bad User
encrypt passwords = yes
pam password change = no
null passwords = yes
force directory mode = 0777
force create mode = 0777
max connections = 5
socket options = IPTOS_LOWDELAY TCP_NODELAY SO_KEEPALIVE
obey pam restrictions = no
use spnego = yes
client use spnego = no
disable spoolss = yes
host msdfs = no
strict allocate = no
wide links = no
bind interfaces only = yes
interfaces = lo br0 192.168.1.1/255.255.255.0
use sendfile = yes
map archive = no
map hidden = no
map read only = no
map system = no
store dos attributes = no
dos filemode = yes
oplocks = yes
level2 oplocks = yes
kernel oplocks = no
wins support = yes
os level = 255
domain master = yes
local master = yes
preferred master = yes
enable core files = no
deadtime = 30
load printers = no
printable = no
max protocol = SMB2
smb encrypt = disabled
min receivefile size = 16384
passdb backend = smbpasswd
smb passwd file = /etc/samba/smbpasswd
Any suggestions as to how to address this?