bringing up a new hme interface

Brian Seppanen

unread,

Mar 25, 2002, 8:13:12 AM3/25/02

to

I have an E250 that has two interfaces. dmesg shows that both
interfaces are initialized at boot. My problem is doing an

#ifconfig hme1 inet 192.168.0.1 netmask 255.255.255.0 broadcast
255.255.255.255

gives the following error

ifconfig: SIOCGLIFNETMASK: hme1: no such interface

is my syntax wrong? Or is there prep work that needs to be done in
advance of bringing up an interface that hasn't been used previously.
It has been cabled up and should be ready to pass traffic, once I bring
it online.

Thanks,

Brian Seppanen
se...@chartermi.net

John Davis

unread,

Mar 25, 2002, 3:04:06 PM3/25/02

to

did you run

# ifconfig hme1 plumb

You can also create an /etc/hosthame.hme1 file with a valid hostname
in it, and see what happens during bootup

Brian Seppanen <se...@chartermi.net> wrote in message news:<3C9F2268...@chartermi.net>...

Brian Seppanen

unread,

Mar 26, 2002, 4:15:51 PM3/26/02

to

Can you run ifconfig hme1 plumb on a running box?

I want to add the interface without taking the box down if possible.

Thanks,

Joe Bloggs

unread,

Mar 26, 2002, 5:06:25 PM3/26/02

to

"Brian Seppanen" <se...@chartermi.net> wrote in message

news:3CA0E507...@chartermi.net...

> Can you run ifconfig hme1 plumb on a running box?
>
> I want to add the interface without taking the box down if possible.
>

Yep.

Brian Seppanen

unread,

Mar 27, 2002, 8:31:11 AM3/27/02

to

So I create a /etc/hostname.hme1 entry and a matching entity in
/etc/inet/hosts

I try ifconfig hme1 plumb and I get the following error

ifconfig: SIOCSLIFNAME for ip: hme1: no such interface

I'm no where near the box but I've had someone visually confirm that the
cabling is complete on the box and that there is a second interface.

I now see the following messages in dmesg

[ID 801518 kern.notice] ip_rput_dlpi(hme1): DL_ERROR_ACK for DL_A
TTACH_REQ(11), errno 8, unix 0

These boot messages also seem to indicate that there is a second
interface on this box.

Mar 11 14:11:31 host hme: [ID 517527 kern.info] SUNW,hme1 : PCI IO 2.0
(Rev Id = c1) Found
Mar 11 14:11:31 host hme: [ID 517527 kern.info] SUNW,hme1 : Local
Ethernet address = 8:0:2
0:cf:29:23
Mar 11 14:11:31 host hme: [ID 517527 kern.info] SUNW,hme1 : Using local
MAC address
Mar 11 14:11:31 host pci_pci: [ID 370704 kern.info] PCI-device:
SUNW,hme@0,1, hme1
Mar 11 14:11:31 host genunix: [ID 936769 kern.info] hme1 is
/pci@1f,4000/pci@2/SUNW,hme@0,
1

I don't think I stated this previously and that was stupid, but this is
a Solaris 8 host. It is a Sun E250 (sparc).

Thanks for all of the help. I appreciate it.

Later,

John Davis

unread,

Mar 27, 2002, 1:49:20 PM3/27/02

to

what does

# /usr/platform/sun4u/sbin/prtdiag -v

say? this is strange, if you see it in dmesg, it should be working fine.

Brian Seppanen <se...@chartermi.net> wrote in message news:<3CA1C99F...@chartermi.net>...

Dave Pyles

unread,

Mar 27, 2002, 4:46:28 PM3/27/02

to

Try this:
grep hme /etc/path_to_inst
Do you see something like this:
"/sbus@7,0/SUNW,hme@3,8c00000" 1 "hme"
"/sbus@3,0/SUNW,hme@3,8c00000" 0 "hme"
If you only see the "0", have you done a reboot -- -r?
If you have and you only see the one line, your card is not working....

Brian Seppanen

unread,

Mar 28, 2002, 8:24:05 AM3/28/02

to

This is what I currently get.

"/pci@1f,4000/pci@2/SUNW,hme@0,1" 1 "hme"
"/pci@1f,4000/network@1,1" 0 "hme"

I have not rebooted the box yet.

Dave Pyles wrote:

> Try this:
> grep hme /etc/path_to_inst
> Do you see something like this:
> "/sbus@7,0/SUNW,hme@3,8c00000" 1 "hme"
> "/sbus@3,0/SUNW,hme@3,8c00000" 0 "hme"
> If you only see the "0", have you done a reboot -- -r?
> If you have and you only see the one line, your card is not working....
>
> Brian Seppanen wrote:
>

Thanks,

Brian Seppanen
se...@chartermi.net

Brian Seppanen

unread,

Mar 28, 2002, 8:30:24 AM3/28/02

to

> Just a SWAG, but you rebooted the system with -r to reconfigure the
> hardware, right?

No I haven't, but I betting that will resolve the problem. The drives
were taken from another E250 and shipped to be put in this box. I bet
a reconfigure will get it. Although the other E250 had a second
interface as well, but oh well.

Sorry to take up everyone's time.

Thanks,
Brian Seppanen
se...@chartermi.net

Brian Seppanen

unread,

Mar 28, 2002, 8:27:19 AM3/28/02

to

John Davis wrote:

> what does
>
> # /usr/platform/sun4u/sbin/prtdiag -v
>
> say? this is strange, if you see it in dmesg, it should be working fine.
>

There is nothing in the prtdiag output regarding any pci cards other
than the video card...

========================= IO Cards =========================

Bus Freq
Brd Type MHz Slot Name Model
--- ---- ---- ---- -------------------------------- ----------------------
SYS PCI 33 3 TSI,gfxp GFXP

No failures found in System
===========================

Thanks,

Brian Seppanen
se...@chartermi.net

Darren Dunham

unread,

Mar 28, 2002, 11:33:16 AM3/28/02

to

In comp.unix.solaris Brian Seppanen <se...@chartermi.net> wrote:
>> Just a SWAG, but you rebooted the system with -r to reconfigure the
>> hardware, right?

> No I haven't, but I betting that will resolve the problem. The drives
> were taken from another E250 and shipped to be put in this box. I bet
> a reconfigure will get it. Although the other E250 had a second
> interface as well, but oh well.

Boot -r (reconfig) boots are not necessary for interfaces. A simple
boot after installing the hardware should be sufficient.

When you reconfig after adding a drive, the sd driver add nice device
links for you (like /dev/rdsk/c0t0d0s0).

There are no instance specific links for interfaces. There's just one
for the driver (like /dev/hme). Boot -r isn't necessary for using
additional interfaces on the same driver.

Let us know if the reboot works, but I'll be surprised if it does.

--
Darren Dunham ddu...@taos.com
Unix System Administrator Taos - The SysAdmin Company
Got some Dr Pepper? San Francisco, CA bay area
< How are you gentlemen!! Take off every '.SIG'!! >