Hanging Login prompt

[terrywashington]

We have an E450 that is plugged into a hub running at 10Mbps half-duplex. When logging in remotely there is a long delay before a login prompt is presented. The server has a second interface that is not used. Below are the lines that I have added to /etc/system to force it to 10Mbps half-duplex. I have also included the results of "ndd -get" and "netstat -rn" and dmesg. If I interpret the results of the ndd -get correctly it indicates that the interface is running at 100Mbps full-duplex while dmesg indicates 10Mbps half-duplex. Could this be the problem?

http://wwwserver3#

more /etc/system |grep hme
set hme:hme_adv_autoneg_cap=0
set hme:hme_adv_100fdx_cap=0
set hme:hme_adv_100hdx_cap=0
set hme:hme_adv_10fdx_cap=0
set hme:hme_adv_10hdx_cap=1

http://wwwserver3#

ndd -get /dev/hme link_status
1

http://wwwserver3#

ndd -get /dev/hme link_speed
1

http://wwwserver3#

ndd -get /dev/hme link_mode
1

http://wwwserver3#

http://wwwserver3#

netstat -rn

Routing Table: IPv4
Destination Gateway Flags Ref Use Interface
-------------------- -------------------- ----- ----- ------ ---------
10.10.168.0 10.10.168.137 U 1 2 hme0
224.0.0.0 10.10.168.137 U 1 0 hme0
default 10.10.168.10 UG 1 2
127.0.0.1 127.0.0.1 UH 2 6 lo0

http://wwwserver3#

dmesg |grep hme
Apr 2 13:12:46

http://wwwserver3

hme: [ID 517527 kern.info] SUNW,hme1 : PCI IO 2.0 (Rev Id = c1) Found
Apr 2 13:12:46

http://wwwserver3

hme: [ID 517527 kern.info] SUNW,hme1 : Local Ethernet address = 8:0:20:8a:0:92
Apr 2 13:12:46

http://wwwserver3

pci_pci: [ID 370704 kern.info] PCI-device: SUNW,hme@0,1, hme1
Apr 2 13:12:46

http://wwwserver3

genunix: [ID 936769 kern.info] hme1 is /pci@1f,4000/pci@4/SUNW,hme@0,1
Apr 2 13:12:51

http://wwwserver3

hme: [ID 517527 kern.info] SUNW,hme0 : Internal Transceiver Selected.
Apr 2 13:12:51

http://wwwserver3

hme: [ID 517527 kern.info] SUNW,hme0 : 10 Mbps Half-Duplex Link Up

http://wwwserver3#

 

[terrywashington]

One of our network engineers verified that the LAN connection is 10Mbps half-duplex.

 

[Igaduma]

Hi,

in a post from Bluedevil, in reply to messagethread "NIC SPEED" he/she said that the hme initalization order has to be respected in your /etc/system file,
so,

You should place the auto_neg_cap as last line,

100fdx
100hdx
10fdx
10hdx
Autoneg


Cheers,

I

 

[terrywashington]

The lines in /etc/system are now listed as below and the login still hangs. Does anyone have any other ideas?

set hme:hme_adv_100fdx_cap=0
set hme:hme_adv_100hdx_cap=0
set hme:hme_adv_10fdx_cap=0
set hme:hme_adv_10hdx_cap=1
set hme:hme_adv_autoneg_cap=0

 

[bdfry]

we have had a similiar situation like yours and it was related to DNS. We would execute a telnet to a remote machine and receive the: Trying xxx.xxx.xxx.xxx and Connected to host@domainname messages, but would have to wait a long time before the login: prompt would appear. As I said, it was related to DNS issues, so once they were resolved, everything worked w/o error or delay. Hope taht helps in some way. Brantley Fry
Systems Administrator I

 

[bluedevils]

actually tdub's /etc/system originally was the right order.

My advise was even more important when configuring the interface on the fly. If you are remotely terminalled while changing from modes you would be locked out until you physically got to the head of the box. I would tend to believe the ndd over dmesg.

Are you still having troubles? So if I read the post correctly, the network engineer said that the server is running at 10half. He didn't say the network is running at 10half and the server *should* match that.

 

[bluedevils]

*male* unix admin