infra: stuttering red0 connection #86

Closed
opened 2024-03-11 03:45:19 +00:00 by ayakael · 6 comments
ayakael commented 2024-03-11 03:45:19 +00:00 (Migrated from lab.ilot.io)

The i225-V intel NIC sometimes craps out:

[346499.658342] igc 0000:01:00.0 red0: NIC Link is Down
[346511.033545] igc 0000:01:00.0 red0: NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX
[346511.582315] igc 0000:01:00.0 red0: NIC Link is Down
[346526.353579] igc 0000:01:00.0 red0: NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX
[346526.486201] igc 0000:01:00.0 red0: NIC Link is Down
[346528.054184] igc 0000:01:00.0 red0: Register Dump
[346528.054188] igc 0000:01:00.0 red0: Register Name   Value
[346528.054190] igc 0000:01:00.0 red0: CTRL            081c0641
[346528.054193] igc 0000:01:00.0 red0: STATUS          40280681
[346528.054196] igc 0000:01:00.0 red0: CTRL_EXT        100000c0
[346528.054198] igc 0000:01:00.0 red0: MDIC            18017949
[346528.054200] igc 0000:01:00.0 red0: ICR             00000001
[346528.054202] igc 0000:01:00.0 red0: RCTL            04408022
[346528.054207] igc 0000:01:00.0 red0: RDLEN[0-3]      00001000 00001000 00001000 00001000
[346528.054211] igc 0000:01:00.0 red0: RDH[0-3]        00000079 00000000 00000000 00000000
[346528.054216] igc 0000:01:00.0 red0: RDT[0-3]        00000078 000000ff 000000ff 000000ff
[346528.054221] igc 0000:01:00.0 red0: RXDCTL[0-3]     02040808 02040808 02040808 02040808
[346528.054226] igc 0000:01:00.0 red0: RDBAL[0-3]      ffffb000 ffffa000 ffff9000 ffff8000
[346528.054231] igc 0000:01:00.0 red0: RDBAH[0-3]      00000000 00000000 00000000 00000000
[346528.054233] igc 0000:01:00.0 red0: TCTL            a503f0fa
[346528.054237] igc 0000:01:00.0 red0: TDBAL[0-3]      fffff000 ffffe000 ffffd000 ffffc000
[346528.054242] igc 0000:01:00.0 red0: TDBAH[0-3]      00000000 00000000 00000000 00000000
[346528.054247] igc 0000:01:00.0 red0: TDLEN[0-3]      00001000 00001000 00001000 00001000
[346528.054251] igc 0000:01:00.0 red0: TDH[0-3]        000000d1 000000ea 0000003b 000000aa
[346528.054256] igc 0000:01:00.0 red0: TDT[0-3]        000000d8 000000f8 0000004c 000000bd
[346528.054260] igc 0000:01:00.0 red0: TXDCTL[0-3]     02100108 02100108 02100108 02100108
[346528.054262] igc 0000:01:00.0 red0: Reset adapter
[346555.702260] igc 0000:01:00.0 red0: NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX
[346556.296017] igc 0000:01:00.0 red0: NIC Link is Down
[346570.624168] igc 0000:01:00.0 red0: NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX
[346571.200929] igc 0000:01:00.0 red0: NIC Link is Down
[346588.993132] igc 0000:01:00.0 red0: NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX
[346589.085824] igc 0000:01:00.0 red0: NIC Link is Down
[346600.424997] igc 0000:01:00.0 red0: NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX
[346601.009756] igc 0000:01:00.0 red0: NIC Link is Down
[346618.762980] igc 0000:01:00.0 red0: NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX
[346618.895651] igc 0000:01:00.0 red0: NIC Link is Down
[346642.230753] igc 0000:01:00.0 red0: NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX
[346642.743535] igc 0000:01:00.0 red0: NIC Link is Down
[346657.117739] igc 0000:01:00.0 red0: NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX
[346657.647514] igc 0000:01:00.0 red0: NIC Link is Down
[346672.464586] igc 0000:01:00.0 red0: NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX
[346672.552407] igc 0000:01:00.0 red0: NIC Link is Down

This looks to be a known issue with these cards. I remember them occurring more often in previous version of Ipfire, but on 6.1 they occurred less often. Next ipfire upgrade will include kernel 6.6 with hopefully even more fixes. Gonna keep this issue up to follow this.

The i225-V intel NIC sometimes craps out: ``` [346499.658342] igc 0000:01:00.0 red0: NIC Link is Down [346511.033545] igc 0000:01:00.0 red0: NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX [346511.582315] igc 0000:01:00.0 red0: NIC Link is Down [346526.353579] igc 0000:01:00.0 red0: NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX [346526.486201] igc 0000:01:00.0 red0: NIC Link is Down [346528.054184] igc 0000:01:00.0 red0: Register Dump [346528.054188] igc 0000:01:00.0 red0: Register Name Value [346528.054190] igc 0000:01:00.0 red0: CTRL 081c0641 [346528.054193] igc 0000:01:00.0 red0: STATUS 40280681 [346528.054196] igc 0000:01:00.0 red0: CTRL_EXT 100000c0 [346528.054198] igc 0000:01:00.0 red0: MDIC 18017949 [346528.054200] igc 0000:01:00.0 red0: ICR 00000001 [346528.054202] igc 0000:01:00.0 red0: RCTL 04408022 [346528.054207] igc 0000:01:00.0 red0: RDLEN[0-3] 00001000 00001000 00001000 00001000 [346528.054211] igc 0000:01:00.0 red0: RDH[0-3] 00000079 00000000 00000000 00000000 [346528.054216] igc 0000:01:00.0 red0: RDT[0-3] 00000078 000000ff 000000ff 000000ff [346528.054221] igc 0000:01:00.0 red0: RXDCTL[0-3] 02040808 02040808 02040808 02040808 [346528.054226] igc 0000:01:00.0 red0: RDBAL[0-3] ffffb000 ffffa000 ffff9000 ffff8000 [346528.054231] igc 0000:01:00.0 red0: RDBAH[0-3] 00000000 00000000 00000000 00000000 [346528.054233] igc 0000:01:00.0 red0: TCTL a503f0fa [346528.054237] igc 0000:01:00.0 red0: TDBAL[0-3] fffff000 ffffe000 ffffd000 ffffc000 [346528.054242] igc 0000:01:00.0 red0: TDBAH[0-3] 00000000 00000000 00000000 00000000 [346528.054247] igc 0000:01:00.0 red0: TDLEN[0-3] 00001000 00001000 00001000 00001000 [346528.054251] igc 0000:01:00.0 red0: TDH[0-3] 000000d1 000000ea 0000003b 000000aa [346528.054256] igc 0000:01:00.0 red0: TDT[0-3] 000000d8 000000f8 0000004c 000000bd [346528.054260] igc 0000:01:00.0 red0: TXDCTL[0-3] 02100108 02100108 02100108 02100108 [346528.054262] igc 0000:01:00.0 red0: Reset adapter [346555.702260] igc 0000:01:00.0 red0: NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX [346556.296017] igc 0000:01:00.0 red0: NIC Link is Down [346570.624168] igc 0000:01:00.0 red0: NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX [346571.200929] igc 0000:01:00.0 red0: NIC Link is Down [346588.993132] igc 0000:01:00.0 red0: NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX [346589.085824] igc 0000:01:00.0 red0: NIC Link is Down [346600.424997] igc 0000:01:00.0 red0: NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX [346601.009756] igc 0000:01:00.0 red0: NIC Link is Down [346618.762980] igc 0000:01:00.0 red0: NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX [346618.895651] igc 0000:01:00.0 red0: NIC Link is Down [346642.230753] igc 0000:01:00.0 red0: NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX [346642.743535] igc 0000:01:00.0 red0: NIC Link is Down [346657.117739] igc 0000:01:00.0 red0: NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX [346657.647514] igc 0000:01:00.0 red0: NIC Link is Down [346672.464586] igc 0000:01:00.0 red0: NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX [346672.552407] igc 0000:01:00.0 red0: NIC Link is Down ``` This looks to be a known issue with these cards. I remember them occurring more often in previous version of Ipfire, but on 6.1 they occurred less often. Next ipfire upgrade will include kernel 6.6 with hopefully even more fixes. Gonna keep this issue up to follow this.
ayakael commented 2024-03-11 03:45:28 +00:00 (Migrated from lab.ilot.io)
https://www.tomshardware.com/news/intel-patches-stuttering-ethernet-issues-but-its-just-a-workaround-for-now
ayakael commented 2024-03-11 04:58:45 +00:00 (Migrated from lab.ilot.io)

The issue occurs again even on 6.6. Its interesting that it occurs now that I'm saturating the routing of the server. Anyways, I've attached a 10gbit nic to red0 due to the unstability.

The issue occurs again even on 6.6. Its interesting that it occurs now that I'm saturating the routing of the server. Anyways, I've attached a 10gbit nic to red0 due to the unstability.
ayakael commented 2024-03-11 16:14:05 +00:00 (Migrated from lab.ilot.io)

We might just skip this and eventually upgrade to http://www.iocrest.com/index.php?id=2396 which supports up to 10gbit. Right now the setup is fine with 1gibt for red0 given that the internet is 1gbit as well. I'll first try the other 2.5gbit card I have at the next maintenance window.

We might just skip this and eventually upgrade to http://www.iocrest.com/index.php?id=2396 which supports up to 10gbit. Right now the setup is fine with 1gibt for red0 given that the internet is 1gbit as well. I'll first try the other 2.5gbit card I have at the next maintenance window.
ayakael commented 2024-04-13 23:46:16 +00:00 (Migrated from lab.ilot.io)

Since the maintenance window, red0 as stuttered a handful of times:

Apr  7 13:45:06 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Down
Apr  7 13:45:10 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Up 2500 Mbps Full Duplex, Flow Control: RX
Apr  8 10:21:22 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Down
Apr  8 10:21:26 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Up 2500 Mbps Full Duplex, Flow Control: RX
Apr  9 08:18:44 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Down
Apr  9 08:18:48 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Up 2500 Mbps Full Duplex, Flow Control: RX
Apr  9 11:45:28 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Down
Apr  9 11:45:32 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Up 2500 Mbps Full Duplex, Flow Control: RX
Apr 10 07:36:23 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Down
Apr 10 07:36:27 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Up 2500 Mbps Full Duplex, Flow Control: RX
Apr 11 06:01:35 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Down
Apr 11 06:01:39 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Up 2500 Mbps Full Duplex, Flow Control: RX
Apr 13 01:28:41 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Down
Apr 13 01:28:45 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Up 2500 Mbps Full Duplex, Flow Control: RX
Apr 13 12:00:46 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Down
Apr 13 12:00:50 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Up 2500 Mbps Full Duplex, Flow Control: RX

I'm still looking for the 2nd NIC we had.

Since the maintenance window, red0 as stuttered a handful of times: ``` Apr 7 13:45:06 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Down Apr 7 13:45:10 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Up 2500 Mbps Full Duplex, Flow Control: RX Apr 8 10:21:22 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Down Apr 8 10:21:26 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Up 2500 Mbps Full Duplex, Flow Control: RX Apr 9 08:18:44 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Down Apr 9 08:18:48 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Up 2500 Mbps Full Duplex, Flow Control: RX Apr 9 11:45:28 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Down Apr 9 11:45:32 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Up 2500 Mbps Full Duplex, Flow Control: RX Apr 10 07:36:23 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Down Apr 10 07:36:27 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Up 2500 Mbps Full Duplex, Flow Control: RX Apr 11 06:01:35 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Down Apr 11 06:01:39 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Up 2500 Mbps Full Duplex, Flow Control: RX Apr 13 01:28:41 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Down Apr 13 01:28:45 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Up 2500 Mbps Full Duplex, Flow Control: RX Apr 13 12:00:46 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Down Apr 13 12:00:50 artalus kernel: igc 0000:01:00.0 red0: NIC Link is Up 2500 Mbps Full Duplex, Flow Control: RX ``` I'm still looking for the 2nd NIC we had.
ayakael commented 2024-06-07 19:18:58 +00:00 (Migrated from lab.ilot.io)

mentioned in issue #138

mentioned in issue #138
ayakael added this to the Infrastructure project 2024-08-25 17:18:11 +00:00
Owner

Closing as i225-V is a lost cause. Everything is off of the 10gbit NIC now

Closing as i225-V is a lost cause. Everything is off of the 10gbit NIC now
Sign in to join this conversation.
No milestone
No project
No assignees
2 participants
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference: ilot/issues#86
No description provided.