I have a SUnfire V120 solaris 10 machine I have recently added memory to. After installation of the RAM, powering back on, the machine did not come back on the network. Consoled into the machine, the routing table has been wiped out completely and I had to reconfigure the IP address. I got the IP address reconfigured, but I cannot add a default gateway. I keep getting network unreachable. I am unable to add routes back into the routing table.
I have changed the network cable. I have even tried on the eri1 interface (it's main configuration was on the eri0). I still got the same result.
So, I am thinking perhaps the hardware for the ethernet ports have failed on the shutdown/reboot. Is there something else I can try to get my server back online? or should I find a new server or a Solaris specific tech that can possibly troubleshoot and replace the parts that are broken?
The memory is reading find. I don't think the RAM has affected it. I just think the server is old and hardware failure. I would like to exhaust all my options before officially declaring it "dead". 
Can you post output of ifconfig -a after you have configured the IP addresses for interfaces?
# ifconfig -a
lo0: flags=1000849<UP,LOOPBACK,RUNNING,MULTICAST,IPv4> mtu 8232 index 1
inet 127.0.0.1 netmask ff000000
eri0: flags=1000851<UP,POINTOPOINT,RUNNING,MULTICAST,IPv4> mtu 1500 index 2
inet 192.168.209.83 --> 192.168.209.83 netmask ffffff00
ether 0:3:ba:2d:54:6a
---------- Post updated at 11:01 AM ---------- Previous update was at 10:48 AM ----------
This is what happens when i restart the network interface
# init.d/network restart
configuring IPv4 interfaces: eri0.
Hostname: unknown
#
hostname file is valid and working
# more hostname.eri0
192.168.209.83 lvccirt03
#
So i'm not sure why it's saying hostname unknown either. Definitely a strange new problem for me. Typcially resetting the network has fixed in the past. But nothing seems to be working.
After you assign 192.168.209.83 as IP address, can you ping gateway?
From the out's it is showing as POINTOPOINT, so you can't expect it to be broadcasting on the network. I was in similar situation(where network was not detected, after I added additional eth's interface on a VM image)
So first check through /var/log/messages about any info on this scenario
Or simply go for reconfiguring the eri0 mate.
check through dladm show-dev for available interface..
This is b/c you are configuring the eri0 as pointopoint, check the flags:
<UP,POINTOPOINT,RUNNING,MULTICAST,IPv4>
Take a look at the configuration files,
If you want to do the configuration manually you can still do:
ifconfig eri0 unplumb
ifconfig eri0 plumb
ifconfig er0 <IP> netmask <MASK>
route add default <GW>
and it should work.
Cheers
Juan
[FONT=monospace]
I cannot ping the gateway. if I try to add the gateway:
# route add default 192.168.209.1
add net default: gateway 192.168.209.1: Network is unreachable
# ping 192.168.209.1
ICMP Host Unreachable from gateway lvccirt03 (192.168.209.83)
for icmp from lvccirt03 (192.168.209.83) to 192.168.209.1
ICMP Host Unreachable from gateway lvccirt03 (192.168.209.83)
for icmp from lvccirt03 (192.168.209.83) to 192.168.209.1
ICMP Host Unreachable from gateway lvccirt03 (192.168.209.83)
and for some screwed up reason, dladm is not installed on this system. I cannot use that. AND there is no /var/log/messages which i find very unusual. i'm fairly certain that pathway was there previously and on the servers it is identical to. There is definitely something unusual going on with this machine.
The error is in the hostname.eri0
It should only contain the name of the interface, in your case:
lvccirt03
But now you have
192.168.209.83 lvccirt03
Remove the IP.
Make sure you have this entry in the /etc/hosts:
192.168.209.83 lvccirt03
THen restart networking, that should fix your issue.
Make sure you understand how to configure networking on solaris 10, you can find tutorials on the web .
Good luck
Juan
1 Like
Ok, Juan!
I deleted the IP for the hostname.eri0 file. I restarted network services... and I was able to add the default route! BUT I still can't ping the gateway.
# /etc/init.d/network restart
configuring IPv4 interfaces: eri0.
Hostname: unknown
# ping lvccirt03
lvccirt03 is alive
# route add default 192.168.209.1
add net default: gateway 192.168.209.1
# ping 192.168.209.1
no answer from 192.168.209.1
# ifconfig -a
lo0: flags=1000849<UP,LOOPBACK,RUNNING,MULTICAST,IPv4> mtu 8232 index 1
inet 127.0.0.1 netmask ff000000
eri0: flags=1000843<UP,BROADCAST,RUNNING,MULTICAST,IPv4> mtu 1500 index 3
inet 192.168.209.83 netmask ffffff00 broadcast 192.168.209.255
ether 0:3:ba:2d:54:6a
#
So it's a little progress! 
I can apparently add routes now. just still can't get out on the network.
if you cannot see the gateway then:
- the network information you have is wrong (let say IP, Netmask, GW)
- Cable problem
- Switch problem, some switches shut the port down if they detect errors.
- NIC problem
- Might be that the GW is configured to not respond to ping as well, have you tried to ping another box on the same network, and on another network?
If the network information is ok... first thing I would try is take a laptop, configure the same IP information , plug the network cable and try.
Oddly I can ping out to all the other addresses on the routing table I added but I cannot ping the gateway. BUT connetivity has returned as I can get to the server remotely now! So I think that fixed it. Everything seems okay now. I will monitor it for the next couple days. Luckily, I don't need to add any m ore hardware to it! so I won't be shutting it down again.
Thank you everyone!
---------- Post updated at 02:50 PM ---------- Previous update was at 02:49 PM ----------
It is possible that the gateway is configured not to respond to pings. Some of our switches/fw/servers are configured to block ICMP. So it's plausible.
network is unreachable means .. you didt assign proper gateway address .... check once what is the gateway .....
you assign ip address in eri1 interface and add gateway ... then you ping
Sys logs on solaris: /var/adm/messages