I'm getting the following errors every few hours:
daemon.err nlbwmon: Netlink receive failure: Out of memory
daemon.err nlbwmon: Unable to dump conntrack: No buffer space available
I've already increased buffer size to 10M but still getting these errors. Also changed the Commit interval to 1hr, hoping that this commit will flush the buffer after writing to the database.
/etc/config/nlbwmon: option netlink_buffer_size 10485760
# cat /proc/sys/net/netfilter/nf_conntrack_count
# cat /proc/sys/net/netfilter/nf_conntrack_max
Any suggestions on fixing this? Thanks.
decreasing the buffer size ?
Thanks for the response.... but not sure I understand your suggestion for decreasing the buffer size. Is this because you're thinking I'm running out of system RAM? I'm currently running on RPi4 with 2GB RAM and I'm only using 6% 115MB.
So no regular router ...
In that case, no, there's no point of shrinking the buffer.
Make this something like 20 mins? Just a wild guess though. Also on RPi try making the buffer say 100MB you have the ram.
Increasing more buffer will be my next try. Didnt think 1 hr of data will take up that much buffer. Saw another user with 3x my conntrack count fixed this same error with only 1M buffer, so thinking 10x will fix my issue but guess not.
Update: Unfortunately, still getting same error after increasing buffer to 100M and restarting nlbwmon. Strange thing is, even after increasing the buffer to 100M and running for a few hrs and getting this error, my total memory usage still shows the same at 115MB. I'd think it will increase by 100M? I've logged this as an issue with the developer as well. If anyone has other suggestions.... Thanks.
same issue still happening on the latest nightly build as of today. 2021-02-15.
I set my netlink buffer to 8MB, 16MB, 32MB, 48MB, and 64MB and still run into the error. I set the commit interval to 40s and refresh interval to 20s. Everything else is default.
I logged this as an issue with the developer 2 mths ago but didnt even get any response from him. Github issue is still in open status. So I just uninstalled the pkg and not use it!
I am seeing a similar issue and noticed that
netlink_buffer_size is used by
setsockopt to set
SO_RCVBUF, but that value cannot exceed
net.core.rmem_max. Quick search shows that there is no error returned if the size exceeds
rmem_max. If anyone is still having this issues, try increasing
net.core.rmem_max ro match
The maximum sizes for socket buffers declared via the SO_SNDBUF
and SO_RCVBUF mechanisms are limited by the values in the
/proc/sys/net/core/rmem_max and /proc/sys/net/core/wmem_max files.
Note that TCP actually allocates twice the size of the buffer requested
in the setsockopt(2) call, and so a succeeding getsockopt(2) call will
not return the same size of buffer as requested in the setsockopt(2) call.
TCP uses the extra space for administrative purposes and internal
kernel structures, and the /proc file values
reflect the larger sizes compared to the actual TCP windows.
On individual connections, the socket buffer size must be set prior to
the listen(2) or connect(2) calls in order to have it take effect.
See socket(7) for more information.