asked on

Weird 2008 to 2008 Issues

This is weird, i am looking at a network i have been asked to clean up after the person responsible for it walked.

There were DNS issues amongst other things, but now this has me a little stumped...

The site has:

2x 2008 AD Server
1x 2003 AD Server
1x 2000 AD Server (Yeah I know)

Alongside multiple 2003 & 2008 Member servers.

I have a weird issues were a 2008 AD Server running Backup Exec cannot talk to another 2008 Member server running SQL for backups.

Neither server can, UNC or Ping each other, yet all other servers and PCs can access both of them. It seems like there is a wall between these two server, but I cant work out why.

There are no errors in the event logs and the Firewall has been disabled by my predecessor.

Can anyone offer any clues?

Juan Ocasio

Are they on the same VLan?

Auratek

ASKER

They should be they are both on the same IP range of 10.0.0.0/24 along with the other servers.

For some reason the SQL server has a second IPv4 assigned to a single LAN adapter.

Juan Ocasio

I'd also check to see if there are any access lists or some other rules preventing them from talking with each other. My first theory on the Vlan is probably incorrect as all other pcs can ping both.

jocasio

Juan Ocasio

Is that second NIC plugged in? If so, is it plugged into the same switch? What kind of switch are they using?

Juan Ocasio

BTW: When you try to ping the box, what is the response?

Auratek

ASKER

The second nic had a 169 address so its now disabled.

for some reason there is an alt ip on the first nic it responds to both 10.0.0.30 and 201.
Not sure on the switches yet but I think you might be right.

Juan Ocasio

Did you disable the second one? a 169 address doesn't mean it's disabled, just means it can't reach a DHCP server. Are the servers pointing to the same DNS Server?

Auratek

ASKER

lan 2 disabled.
there are 3 dns servers .6 .104 & .105
sql looks to 6 & 104
the ad server looks to itself (105) and 104

Auratek

ASKER

oh the ad server is physical whilst sql is a vmware server

Juan Ocasio

Do you know why do they have 3 DNS servers? Did you try to do a NSLookup on each box to see what host name they have associated with each IP address? There must be something goofy going on where the packets can't be delivered because either the destination is unknown and they're being dropped, or because they're being refused. Seems weird to have 3 DNS Servers for such a small company (I'm assuming small because of the IP range size)

Santosh Gupta

hi,

1. are you able to ping with IP ?
2. pls share the ping and nslookup result.
3. run "arp -a ipofanyserver" and share the result.
4. are they connected to same switch, if you are using managed switch then pls check the switch configuration.

Auratek

ASKER

looking at the switch now. its a dell device setup but what seems to be a moron as there and so many vlans. found out that vms on 1 server are on 1 vlan and vms on another server are on another vlan.

its s mess im now thinking about tell them to start again

Juan Ocasio

There's your problem. You have to get the ports on the same VLAN. what you can do to test it is plug the server into another port on the same vlan as the other server and then try to ping it. They were probably trying to separate access based on needs, but somehow screwed it all up.

Auratek

ASKER

Let give these some names

AD1 cannot talk to SQL1
AD1 can talk to AD2
AD2 can talk to SQL
AD1 can talk to member1
Member 1 can talk to SQL1

AD1 is on an unknown Vlan
AD2 is on Vlan7
SQL1 is on Vlan7
Member1 is on Vlan2

All servers can talk to each other no matter which Vlan they are on, its just AD1 & SQL1 that cant talk to each other.

Its rather bizarre and a very messy setup especially since all the servers are on 1 site. Unfortunately I am not at that site and have no way to get hands on with the switch.

I can tell you that the Core switch is a Dell PowerConnect 6248P

Santosh Gupta

move the AD1 to VLAN7

Juan Ocasio

I would strip away the vlans (or get an unmanaged switch and plug everything in) to see if that resolves the problem. You never mentioned what the response was for the ping from AD1 to SQL1. Does it timeout or is the host unreachable?

Auratek

ASKER

Sorry its destination host unreachable.

Yeah i think its more infrastructure than windows too, so i think ill call it a day on this until i can get hands on with the core switch.

Juan Ocasio

There may be an issue with the default gateway on AD1. Host Unreachable means the packets aren't being routed to the proper destination and being dropped. Is the default gateway the same as the other servers?

Auratek

ASKER

Yeah the default gateway on all devices is 10.0.0.10. The core Switch.

I have looked over the switch and cannot find any entries to say why 10.0.0.105 cannot talk to 10.0.0.30.

The Data Vlan is 7 and i can access devices on that Vlan from AD1, just no the SQL server on it.

Juan Ocasio

how many ports on the switch? I would get an unmanaged switch and plug everything in to see if that resolves your problem.

ASKER CERTIFIED SOLUTION

Auratek

membership

This solution is only available to members.

To access this solution, you must be a member of Experts Exchange.

Start Free Trial

Auratek

ASKER

fixed