Netdata 10min_fifo_errors alerts

[thed3v]

Hello!

Since i've installed Netdata app i'm getting a lot of 10min_fifo_errors alerts when there is higher network traffic, like when downloading something.
Judging by the description here: https://community.netdata.cloud/t/10min-fifo-errors/1814 it sounds like there's something misconfigured on my machine - i.e. RX ring buffer size needs to be increased? I'm not really sure. Has anyone else experienced this before?

 

[danwestness]

I am seeing the same. Ever find an answer?

 

[depasseg]

I'm seeing this too. Any updates?

 

[depasseg]

I found this and tried changing the rx size as per the troubleshooting section. We'll see what happens.

10min_fifo_errors

10min_fifo_errors OS: Linux Between the IP stack and the Network Interface Controller (NIC) lies the driver queue. This queue is typically implemented as a FIFO ring buffer into the memory space allocated by the driver. The NIC receive frames and place them into memory as skb_buff data...

community.netdata.cloud

Code:

root@truenas[~]# ethtool -g eno1
Ring parameters for eno1:
Pre-set maximums:
RX:             4096
RX Mini:        n/a
RX Jumbo:       n/a
TX:             4096
Current hardware settings:
RX:             512
RX Mini:        n/a
RX Jumbo:       n/a
TX:             512
root@truenas[~]# ethtool -G eno1 rx 4080
root@truenas[~]# ethtool -g eno1
Ring parameters for eno1:
Pre-set maximums:
RX:             4096
RX Mini:        n/a
RX Jumbo:       n/a
TX:             4096
Current hardware settings:
RX:             4080
RX Mini:        n/a
RX Jumbo:       n/a
TX:             512
root@truenas[~]#

 

[tmbnas]

I found this and tried changing the rx size as per the troubleshooting section. We'll see what happens.

10min_fifo_errors

10min_fifo_errors OS: Linux Between the IP stack and the Network Interface Controller (NIC) lies the driver queue. This queue is typically implemented as a FIFO ring buffer into the memory space allocated by the driver. The NIC receive frames and place them into memory as skb_buff data...

community.netdata.cloud

Code:

root@truenas[~]# ethtool -g eno1
Ring parameters for eno1:
Pre-set maximums:
RX:             4096
RX Mini:        n/a
RX Jumbo:       n/a
TX:             4096
Current hardware settings:
RX:             512
RX Mini:        n/a
RX Jumbo:       n/a
TX:             512
root@truenas[~]# ethtool -G eno1 rx 4080
root@truenas[~]# ethtool -g eno1
Ring parameters for eno1:
Pre-set maximums:
RX:             4096
RX Mini:        n/a
RX Jumbo:       n/a
TX:             4096
Current hardware settings:
RX:             4080
RX Mini:        n/a
RX Jumbo:       n/a
TX:             512
root@truenas[~]#

The same warning on my system (Bluefin 22.12.0) with 2 i350 ports in a bond. Both interfaces are at RX/TX 256. I haven't seen this warning on a second Truenas Scale system with broadcom nics (also in a bond). So it also could be a a misconfiguration in my first system or a driver / firmware issue whereas RX 256 seems to be low or?

Does someone know how this ring buffer value can be set permanent and survives Truenas updates, may be an init script in the Settings/Advanced section?

Thanks