Setting up a network for SOlaris

OK, so I bought an SPARC Ultra on ebay. It still has Solaris 9 loaded, so I am using that.

Hooked up to my home network.

IP address is 192.168.0.10 Netmask is set correct.

I can see my router via browser, but I cannot see the internet..

added my router to /etc/defaultroute, so I can ping my DNS servers, where do I set DNS ?

There was no /etc/resolv.conf or named.conf

I created a reolv.conf with DNS IP, domainname and search (www.google.com), then rebooted. No luck

Tips/Pointers.RTFMs apprecited.

Find more posts tagged with

Comments

nipper

OK, trudging along. People are not reading devnet on their day off. I am shocked.

I changed resolv.conf so now nslookup returns the correct information. But that is not being passed to ping or netscape as ping www.interwoven.com fails, with an unknown host. But if I can nslookup to www.interwoven.com

huh ?

tips/pointers/RTFMs appreciated.

Andy

jed

Few things to check for:

Create a file called...
/etc/defaultrouter
...and put the address of your gateway in here

Check /etc/nsswitch.conf and make sure that things are setup for DNS. Example:
hosts: files dns

--
Jed Michnowicz
jedm@sun.com
Content Management Engineering
Sun Microsystems

skip11

Hi,

with the correct setting of the default router you should see something
like this in netstat -rn

default <ip of your router> UG 1 281664

if you can ping it, and have DNS set up correctly, then getting
out to the innernet should be ok, if not, check browser proxy setting.

regards,

R.Barger
Credit Suisse Group
Zurich, Switzerland

nipper

Thanks it was the nsswitch

I am certain I will have more issues as I try to load gcc, apache, etc.

jed

Blastwave will be your friend for common packages:
blastwave.org

--
Jed Michnowicz
jedm@sun.com
Content Management Engineering
Sun Microsystems

nipper

Thanks, I tried blastwave for apache. However it installed OpenLDAP and a couple other require modules, but hosed my
network

I can ping myself but no other computers on the same network.

My config files do not appear to have changed and ifconfig looks the same.

I never used pkg_add before, I hope I can get a list of packages and uninstall with it, if I cannot get it fixed.

nipper

OK, so any more tips/pointers/etc appreciated

I bought an Ultra Enterprise 2 off ebay. It has Solaris 9 loaded, though no docs or meda.
I have not done serious Sun install since the old SunOS days.

When I added Apache my network connection went away. ifconfig says everythin is OK and
my lights are set, but ping goes no where.

I get an error in CDE that DT Messaging is down, check /etc/hosts, /etc/src.sh and /usr/adm/inetd.sec for
changed hostnames. /etc/hosts is fine and I do not have either of the other 2 files, nor any idea what they are for
or what they look like. Any ideas ?

BTW, Apache works great. I took OpenLDAP out of the startup (rc*.d) but no luck.

UPDATE. Ran sys-unconfig and reconfig'ed to no avail.

Thanks

Andy

Edited by nipper on 01/03/06 02:37 PM (server time).

skip11

Hi nipper,

I have to say that's strange that blastwave brought your network down. really there are only a few
pieces to setting up a sun for a static ip. hosts, nsswitch.conf, resolv.conf, defaultrouter, hostname.qfe*
if you have a quad ether card in there, and netmasks - all found in /etc. Where did you see the error
in DT Messaging ? in the boot cycle, or logging into CDE (ugh, CDE - use gnome 2.0) If you saw the errors
at login, edit the passwd file put in a new location for your home directory. Create the new directory. This will
allow CDE to write clean new files when you attempt to login next. I've seen various dot files that CDE uses get
hosed, maybe that will fix the messaging errors. If you have no luck, post the errors and I'll have a look. Also
post the output of netstat -rn, ipconfig -a (run as root, your MAC will be included - cut it out if you don't want us
to see it). you could also run snoop as root and look at the traffic for things like telnets to the defualtrouter,
failed DNS calls, etc.

regards,

R.Barger
Credit Suisse Group
Zurich, Switzerland

nipper

I have to say that's strange that blastwave brought your network down. really there are only a few
pieces to setting up a sun for a static ip. hosts, nsswitch.conf, resolv.conf, defaultrouter, hostname.qfe*

I certainly agree with that. The files you listed are correct.

if you have a quad ether card in there, and netmasks - all found in /etc. Where did you see the error
in DT Messaging ? in the boot cycle, or logging into CDE (ugh, CDE - use gnome 2.0) If you saw the errors
at login, edit the passwd file put in a new location for your home directory. Create the new directory. This will
allow CDE to write clean new files when you attempt to login next. I've seen various dot files that CDE uses get
hosed, maybe that will fix the messaging errors.

Not certain what a quad card is. I have the built in ethernet that was up. There is also an sbus SCSI/ethernet
card. I configured one then the other (each with sys-unconfig). No luck.

On the Sun forums there is a pretty detailed explanation of this CDE error. Can easily be caused by networking
problems, so I suspect that is a symptom rahter than the issue.

If you have no luck, post the errors and I'll have a look. Also
post the output of netstat -rn, ipconfig -a (run as root, your MAC will be included - cut it out if you don't want us
to see it). you could also run snoop as root and look at the traffic for things like telnets to the defualtrouter,
failed DNS calls, etc.

Since I cannot telnet to it, it is tough, but I think I have seen the issue. Hope you can tell me what to do.

Hostname loki ip 192.168.0.11 defaultrouter 192.168.0.10 direct connect

(I wrote down the output and entered it manually so accuracy is questionable)

ifconfig hme1 (configured 2nd ethernet 1st is not configured, same results the other way around)

hme1: flags=100843<UP,BROADCAST,RUNNING,MULTICAST,IPv4) mtu 1500 index 2
inet 192.168.0.11 netmask FFFFFF00 broadcast 192.168.0.255

netstat -r

Destination Gateway Flag Ref US Interface
192.168.0.0 loki U 1 212 hme1
244.0.0.0 loki U 1 0 hme1
default 192.168.0.0 UG 1 172
localhost localhost UH 42 12486 lo0

You see there is no interface defined for the default route. Is that a problem or does
it then go to the 1st rule which is set on hme1 ?

I will see if I have snoop

Thanks

Andy

skip11

Hi nipper,

first thing I would do is move the main network connection off hme1 and back onto hme0 and reboot.
make sure the defaultrouter file contains the ip of your router: 192.168.0.10. Are you sure this is
corect ? Alot of router products use 192.168.0.1 as the default route. Also, usually in home networks
the router is some sort of DHCP server that can hand out ip's statically as well. config your router to
always hand out the same ip to the MAC of hme0. (I use Netgear, but this function is quite common
in most home use products.) make sure you have /etc/hostname.hme0 - this file contains your hostname
from /etc/hosts, so it should only contain loki. /etc/netmasks should look like this:

192.168.0.0 255.255.255.0

If you have all the necessary files configured correctly and you still cannot
see the defaultrouter, try this:

ifconfig hme0 down
ifconfig hme0 plumb
ifconfig hme0 up

if you still have no connectivity, run the command dmesg and post it.

regards,

R.Barger
Credit Suisse Group
Zurich, Switzerland

nipper

got it

Bad cable, lights were on but no data. Strange, replaced it & I am good

Thanks

Andy