Schrittmotoren brummen im Stillstand

cncfraese@CNCFraese:~$ uname -a
Linux CNCFraese 6.2.0-rt3 #1 SMP PREEMPT_RT Thu Mar 30 17:40:57 CEST 2023 x86_64 GNU/Linux
cncfraese@CNCFraese:~$ ping 10.10.10.10
PING 10.10.10.10 (10.10.10.10) 56(84) bytes of data.
64 bytes from 10.10.10.10: icmp_seq=1 ttl=64 time=0.347 ms
64 bytes from 10.10.10.10: icmp_seq=2 ttl=64 time=0.290 ms
64 bytes from 10.10.10.10: icmp_seq=3 ttl=64 time=0.304 ms
64 bytes from 10.10.10.10: icmp_seq=4 ttl=64 time=0.296 ms
^C
--- 10.10.10.10 ping statistics ---
4 packets transmitted, 4 received, 0% packet loss, time 3077ms
rtt min/avg/max/mdev = 0.290/0.309/0.347/0.022 ms

Julien wrote:

cncfraese@CNCFraese:~$ uname -a
Linux CNCFraese 6.2.0-rt3 #1 SMP PREEMPT_RT Thu Mar 30 17:40:57 CEST 2023 x86_64 GNU/Linux
cncfraese@CNCFraese:~$ ping 10.10.10.10
PING 10.10.10.10 (10.10.10.10) 56(84) bytes of data.
64 bytes from 10.10.10.10: icmp_seq=1 ttl=64 time=0.347 ms
64 bytes from 10.10.10.10: icmp_seq=2 ttl=64 time=0.290 ms
64 bytes from 10.10.10.10: icmp_seq=3 ttl=64 time=0.304 ms
64 bytes from 10.10.10.10: icmp_seq=4 ttl=64 time=0.296 ms
^C
--- 10.10.10.10 ping statistics ---
4 packets transmitted, 4 received, 0% packet loss, time 3077ms
rtt min/avg/max/mdev = 0.290/0.309/0.347/0.022 ms

ping time of 300µs seems way too high, it should be 30µs or less. use dedicated network interface and direct connection to mesa card (no switch), check with ethtool if wirespeed is 100mbps and not 10mbps (could be bad cable).

the Mesa card is connected directly to the Intel NUC 11. No switch in between.
I have changed the cable with one that does my 1000Mbit connection in my home network. Ping remains >300ms.
ethtool output:
Settings for eno1:
Supported ports: [ TP MII ]
Supported link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Full
Supported pause frame use: Symmetric Receive-only
Supports auto-negotiation: Yes
Supported FEC modes: Not reported
Advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Full
Advertised pause frame use: Symmetric Receive-only
Advertised auto-negotiation: Yes
Advertised FEC modes: Not reported
Link partner advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
Link partner advertised pause frame use: Symmetric
Link partner advertised auto-negotiation: Yes
Link partner advertised FEC modes: Not reported
Speed: 100Mb/s
Duplex: Full
Auto-negotiation: on
master-slave cfg: preferred slave
master-slave status: slave
Port: Twisted Pair
PHYAD: 0
Transceiver: external
MDI-X: Unknown
Supports Wake-on: pumbg
Wake-on: d
Link detected: yes

IP of the NUC is 10.10.10.11 netmask 255.255.255.0 static

I have found this topic
forum.linuxcnc.org/27-driver-boards/4691...error-finishing-read
but since this is my first Linux project and I got it for only a couple of days I have no clue what solution they found.

Since you likely have a Intel MAC you should disable IRQ coalescing:

man hm2_eth

Yes, I already did that before.
I double checked it now and did everything in the man including flashing hostmot2 (7i96s_d from your website)
halcmd: loadrt hm2_eth board_ip=192.168.1.121
both successful.

The ping is now around 250ms bit I still get error finishing read! after a few seconds. Sometimes immediately after starting Linuxcnc, sometimes after 10 seconds.

250 usec is pretty bad for a ping

Did you verify that IRQ coalescing is off?

ethtool -c

will show this

If you still have bad Ethernet latency, you may need a different kernel
( many of the default  Debian "real time" kernels aren't )

The latest RT kernel seems quite good (6.3-rc5-rt8)