different vlans

[christina_neo]

Hi,

Would appreciate some advice on the following:

there are no firewall implemented. can we say that network is ok, since ping from serverA to serverB is ok? but nfs is not? does it mean something is wrong in the network? any other things to check?

Hope someone can enlighten.

Thanks.

Christina

[globalnettech]

Hello Christina,

the PING is a basic connectivity check, and tells you that there is basic connectivity between your network devices. However, a successful ping does not necessarily mean that all your applications are reachable, since they use different ports and protocols. What exactly are you running into ?

Regards,

GNT

[zerozerotito]

Hi Christina,

Ping is used to check network connection. If it is succesful, Everything is ok till on Layer 1, 2, and 3.

If you wanna check All layers... try to use telnet... If you can telnet, this means PERFECTO...:)

Hope it helps,

[christina_neo]

Hi GNT,

I am experiencing the following:

server1(e.g.vlan10) and server2(e.g.vlan20) experienced NFS timeout. Starting from server1 to server2, they are 6 hops away. ping is ok.

1) is 6 hops too far away? or

2) should the arp aging time on router and mac address aging time on switches set close to each other?

3) any other suggestions?

Thanks,

christina

[zerozerotito]

Hi Christina

Can you check that interfaces are configured as autoduplex and autospeed by using "sh interfaces"?

If so, please change them.

And Can you tell me what you have between servers?

REgards

[christina_neo]

Hi,

yes, they are configured as auto speed and auto duplex. However, there are no FCS errors, or other indicative errors.

The devices between the servers are switches and routers.

Thanks for your advice.

[zerozerotito]

I think here is the problem... I think you shouldnt use autoduplex and speed values.. Please change them.. Then its gonna be work...

Let me know How it goes.

[Richard Burts]

Christina

As others have pointed out PING is good for testing basic IP connectivity. And you have indicated that ping works, so we can assume that the NFS problem is not a basic connectivity problem.

When I hear that some application do work (PING) and some applications do not work (NFS) one of the first things that I think of is access lists. Is it possible that there is an access list (on either of the routers at the end of the path or on a router in the middle of the path) which is permitting PING but is denying NFS?

HTH

Rick

[christina_neo]

Hi rick,

Btw, the NFS has been running fine for months until recently it has such problems.

I will do a check for the access list. Thanks.

Christina

[mbuillon]

Hi,

In the same LAN you have 6 hop to go from one vlan to one other ?

I thing that only one hop is useful... What is the config of your router ?

Have you got problem with your routing protocole ?

[Richard Burts]

Maxime

I am not sure why you would say that only one hop is useful. In a lab situation perhaps you would only set up one hop but in many live networks it is multiple hops from one part of the network to another part of the network.

As I pointed out in my earlier post if ping works from one device to another then it has demonstrated IP connectivity and it would not be the case that it is a problem with the routing protocol. If it were a problem with the routing protocol then ping would fail.

So we must look for some other explanation of the problem. Christina says that NFS worked ok for months and recently experienced this problem. So I think we should look to find what kind of changes might have been made in the network that would cause this. I have seen a few times symptoms similar to this where NFS experienced problems that were caused by MTU issues when going over a path that re-encapsulated the data packet (over a GRE tunnel or an IPSec VPN or something similar). In addition to my suggestion of looking for access lists it would help to know if anything has changed that would send NFS over an encapsulated hop.

HTH

Rick