xmerlin that disk had a zpool with smartos before? I remember having the same problem and sdc-factory reset help if I'm not mistaken or you could try pxe?

neirac, no the machine is new and the disk is empty

I can boot smartos but there is a strange error pasteboard.co/Ui6NY056aFF8.jpg

xmerlin do you see the network interfaces on the host? dladm show-phys or sysinfo looking at the Network Interfaces tag

I cannot see the network interface and its quite strange ...I'm using this machine with freebsd

xmerlin which nic does the machine has? is supported by SmartOS

intel I226

a revision of i225

xmerlin maybe danmcd has more information or bahamat I don't see that in illumos hcl

i'm 90% sure we don't support that

:(

the auditd is probably related to the network/physical

the logs from network/physical (`svcs -L network/physical` should give you the pathname)

would be good to see

if it's just a revision, i think we have a driver for i225, right?

jbk, pasteboard.co/lx9AMbJBXRQa.jpg and pasteboard.co/8cVaiHSEo1hN.jpg

yeah, so i'm guessing 'dladm show-phys' is probably not showing the interface

I225/I226 are unsupported. We need the igc driver.

oh we don't have it at all.. so more work than just updating the driver

Yes.

i'd love to port it from freebsd, but ENOTIME

Ironically it's actually still an e1000g, but Intel upstream decided on a fresh driver.

(The MAC is the same and the descriptor layout is mostly the same, just not what Intel decided to do)

and probably not something likely to be on systems seen at $WORK

if it's just 1gb

(otherwise I could probably have a case for $WORK to port it as opposed as a personal thing)

It's 2.5G.

jbk: $work still mnx in your case?

no, i've never worked for mnx

jperkin: interesting, I'm hitting this bugzilla.samba.org/show_bug.cgi?id=14685 with smbd 4.18.6

I wondered why dmesg takes so long, turns out /var/adm/messages is 12G in size due to that message :-D

jbk oops, It gets confusing from time to time with people changing $work

rmustacc: IIUC I225 is 2.5, and I226 is 5... right?

Nope... just checked, sorry.

2.5

No, both are 2.5G.

There's not a lot of consumer 5GBASE-T/N right now.

AIUI one can use cat5e for 2.5. I'm not as sure about cat5e for 5. Also AIUI one can *try* to use it for 10, but it's dicey at best.

mine (not quite done yet) has cat6.. just need to find a switch that has at least a couple of 2.5gb ports (for wireless) and a couple of 10gb ports (for storage/file serving)

I've done a lot of 10G over Cat5e in the lab and surprisingly had no problems. Cables I made too.

jbk end user level switch?

jbk I was checking on netgear, well partly because I also got orbi

something for home

not sure how many ports though (need to go back and count the # of drops)

ideally, I'd want every drop plugged into the switch, that way I can just activate/deactive/configure the ports without having to mess with cross connects

probably less of an issue w/ the house, but my experience is not doing that is how you end up with a rat's nest of cables at your patch panels

i also have used 10G over cat5e, seems fine for shortish (~25M) runs

Oooh, 25M is far less than any cable length I have in my house.

Okay, maybe not FAR less (thinks about 1-2 feeds) but not more.

only one of the runs the contactors put it gave me trouble when testing, but that one is just bad even at 10Mbit

if someone(tm) would port igc I could get my hands on hardware for testing

danmcd: Dan, I'll have the Dell PowerEdge 750xs available for testing from tomorrow until end of week. (with that Intel(R) Xeon(R) Silver 4310 CPU ) Any output you'd be interested or just generic works fine/problem with latest smartos would suffice?

This is for ucode update testing?

danmcd: yup, the microcode we talked about here log.omnios.org/smartos/2023-11-14

If you've run pre-ucode PIs, having the output of "ucodeadm -v" would be nice. Then you can compare it to post-ucode PIs.

danmcd: so starting with 20231102T000310Z and upgrading to 20231116T064739Z ?

Yes. Exactly that.

SHould see changes.

I'm out for dinner now. Hope *not* to be back after, but that depends on some tests I'm running.

danmcd: talk to you later and have a nice evening.