topic NIC issue in General Topics

NIC issue

RemoteUser — Mon, 17 Feb 2025 12:14:22 GMT

can someone help me better understand what this output I see in the /var/log/messagges indicates.

is this normal?

kernel:Number of in use tx queues changed invalidating tc mappings. Priority traffic classification disabled!
kernel:ice 0000:4b:00.0 eth5-03: NIC Link is up Unknown bps Full Duplex, Requested FEC: NONE, Negotiated FEC: NONE, Autoneg Advertised: Off, Autoneg Negotiated: False, Flow Control: None

On almost all nterfaces I see this is this normal?
Also this:
kernel:ice 0000:4b:00.0 eth5-03: Please specify at least 1 Rx and 1 Tx channel
kernel:ice 0000:4b:00.3 eth5-02: Please specify at least 1 Rx and 1 Tx channel
kernel:ice 0000:4b:00.1 eth5-01: Please specify at least 1 Rx and 1 Tx channel
kernel:ice 0000:4b:00.2 eth5-04: Please specify at least 1 Rx and 1 Tx channel
kernel:ice 0000:98:00.3 eth6-02: Please specify at least 1 Rx and 1 Tx channel
kernel:ice 0000:98:00.1 eth6-01: Please specify at least 1 Rx and 1 Tx channel
kernel:ice 0000:98:00.0 eth6-05: Please specify at least 1 Rx and 1 Tx channel
kernel:ice 0000:98:00.2 eth6-06: Please specify at least 1 Rx and 1 Tx channel
kernel:ice 0000:98:00.7 eth6-04: Please specify at least 1 Rx and 1 Tx channel
kernel:ice 0000:98:00.6 eth6-08: Please specify at least 1 Rx and 1 Tx channel
kernel:ice 0000:98:00.4 eth6-07: Please specify at least 1 Rx and 1 Tx channel
kernel:ice 0000:98:00.5 eth6-03: Please specify at least 1 Rx and 1 Tx channel
kernel:ice 0000:e3:00.1 eth7-01: Please specify at least 1 Rx and 1 Tx channel
kernel:ice 0000:e3:00.7 eth7-04: Please specify at least 1 Rx and 1 Tx channel
kernel:ice 0000:e3:00.0 eth7-05: Please specify at least 1 Rx and 1 Tx channel
kernel:ice 0000:e3:00.4 eth7-07: Please specify at least 1 Rx and 1 Tx channel
kernel:ice 0000:e3:00.5 eth7-03: Please specify at least 1 Rx and 1 Tx channel
kernel:ice 0000:e3:00.6 eth7-08: Please specify at least 1 Rx and 1 Tx channel
kernel:ice 0000:e3:00.3 eth7-02: Please specify at least 1 Rx and 1 Tx channel

Re: NIC issue

the_rock — Mon, 17 Feb 2025 12:20:31 GMT

Never seen that before. When did this happen?

Andy

Re: NIC issue

RemoteUser — Mon, 17 Feb 2025 12:26:51 GMT

I was looking in /var/log/messages because I have a problem with the Check Point interface. The link status is down—it could be due to a physical connection issue, but I doubt it. So, I was checking /var/log/messages to see if I could find anything useful.
and this one I showed seems to be unusual

Re: NIC issue

the_rock — Mon, 17 Feb 2025 12:29:37 GMT

Right, but you said its the case for almost all interfaces?? How many are showing status as down?

Andy

Re: NIC issue

RemoteUser — Mon, 17 Feb 2025 12:38:31 GMT

I would like to point out that we currently have two interfaces active but we tried to enable one interface connected on the checkpoint but we could see that the ink remained dowmn

Re: NIC issue

Timothy_Hall — Mon, 17 Feb 2025 13:03:44 GMT

Sounds like someone may have messed with the default Multi-Queue configuration which is a big no-no as it interferes with dynamic split and hyperflow. Please post output of mq_mng --show and netstat -ni. All interfaces should report a state of Up, and Mode of Auto or Dynamic. Here is the text warning about this in my Gateway Performance Optimization Course:

Trying to manually change the state or operation of Multi-Queue on individual network interfaces from clish or using the mq_mng command is strongly discouraged under Gaia 3.10, as doing so is likely to disrupt the proper balancing of traffic across the multiple traffic queues thus causing voluminous RX-DRPs and hurting performance, especially if Dynamic Balancing/Split is present.

Examples: sk154392: An available CPU Core is not handling any queue, when using Multi-Q & sk115998: Multi-queue not mapping its queues to CPUs in ascending order & sk173950: High CPU for SND cores due to MQ configuration issues on VSX environment. To determine if this issue is present on your firewall and how to rectify it, see sk168498: High rate of input discards after reboot when Multi-Queue is configured.

The log file for the Multi-Queue feature $FWDIR/log/mq_mng.elg may also be helpful to determine if someone tampered with the Multi-Queue default configuration. Note that if Dynamic Balancing/Split is enabled, you will see a state of "Dynamic" instead of "Auto" as shown in the prior screenshot, thus indicating that Dynamic Balancing/Split has properly taken control of Multi-Queue for that interface. Main Takeaway: Do NOT mess with Multi-Queue, leave the defaults set!

Re: NIC issue

the_rock — Mon, 17 Feb 2025 13:27:40 GMT

One thing you can try is generic linux command to "restart" the interface

ifdown eth0; ifup eth0 (or whatever interface name would be)

Andy

Re: NIC issue

Timothy_Hall — Mon, 17 Feb 2025 13:32:04 GMT

Yes, but be aware that technically this does cause an outage on the interface although it is very brief and normally won't cause a cluster failover if you run it on the active member. One other nice trick is that it also resets the counters shown by netstat -ni for that interface with needing a reboot.

Re: NIC issue

RemoteUser — Mon, 17 Feb 2025 13:34:25 GMT

this is the output:
netstat -ni
Kernel Interface table
Iface MTU Met RX-OK RX-ERR RX-DRP RX-OVR TX-OK TX-ERR TX-DRP TX-OVR Flg
Sync 1500 0 431127888 0 0 0 431683217 0 0 0 BMRU
eth6-01 1500 0 1427315 0 0 0 1458513 0 0 0 BMRU
eth7-01 1500 0 863723 0 0 0 861687 0 0 0 BMRU
eth7-02 1500 0 0 0 0 0 0 0 0 0 BMU
lo 65536 0 4451384 0 0 0 4451384 0 0 0 ALdORU
maas_tunnel 1500 0 281397 0 0 0 451328 0 0 0 MOPRU

Re: NIC issue

RemoteUser — Mon, 17 Feb 2025 16:02:27 GMT

i saw this:

Feb 12 05:31:08 2025 gw-495ded kernel:ice 0000:98:00.1 eth6-01: NIC Link is up Unknown bps Full Duplex, Requested FEC: NONE, Negotiated FEC: NONE, Autoneg Advertised: Off, Autoneg Negotiated: False, Flow Control: None
Feb 12 05:31:08 2025 gw-495ded kernel:Number of in use tx queues changed invalidating tc mappings. Priority traffic classification disabled!
Feb 12 05:31:08 2025 gw-495ded kernel:ice 0000:98:00.0 eth6-05: NIC Link is up Unknown bps Full Duplex, Requested FEC: NONE, Negotiated FEC: NONE, Autoneg Advertised: On, Autoneg Negotiated: False, Flow Control: None
Feb 12 05:31:08 2025 gw-495ded kernel:ice 0000:98:00.0 eth6-05: To change autoneg please use: ethtool -s <dev> autoneg <on|off>

on different interface why? That's indicate that someone change manually the negotiation?

Re: NIC issue

_Val_ — Mon, 17 Feb 2025 19:03:43 GMT

Please open a TAC case for this.

Re: NIC issue

the_rock — Mon, 17 Feb 2025 20:22:12 GMT

Thats what it looks like. here is example from my R82 lab.

Andy

R82> show interface eth0
state on
mac-addr 50:01:00:01:00:00
type ethernet
link-state link up
mtu 1500
auto-negotiation on
speed 1000M
ipv6-autoconfig Not configured
monitor-mode off
duplex full
link-speed 1000M/full
comments external
ipv4-address 172.16.10.253/24
ipv6-address Not Configured
ipv6-local-link-address Not Configured

Statistics:
TX bytes:3284670593 packets:31474311 errors:0 dropped:0 overruns:0 carrier:0
RX bytes:6361787351 packets:32132399 errors:249890 dropped:0 overruns:0 frame:0

SD-WAN: Not Configured
R82>

Re: NIC issue

RemoteUser — Tue, 18 Feb 2025 09:57:52 GMT

Solved, guys! It was because the cable was broken.
Thank all for the support

Re: NIC issue

the_rock — Tue, 18 Feb 2025 12:13:16 GMT

Well, that was an easy fix 🙂

Great job!