Here's a wierd one fer ya

  • Thread starter Thread starter Adysthemic
  • Start date Start date
A

Adysthemic

Group,
I have two servers identical except for drive size.Each was running an
Intel Pro/100 s dual NIC in load sharing configuration.The mainserver
started experiencing connectivity problems.I could ping it,though after the
first
3ms hit it would time out.Re-installing the driver did not help.I replaced
that NIC with a 3com single 90x.Problem solved for the moment.Then I
discover that the other server has exactly the same problem.I swapped that
NIC out for a single 3com.All was well.Any thoughts on what would cause two
identical NICs to fail like that?I cannot experiment until the weekend.All
that mattered at the moment was getting things going again,which I did.
I'm going to re-install the Dual Intels this weekend to see what
happens,but in the mean time,any opinions on this odd event?.Adysthemic
 
You want to configure the dual NICs in failover mode rather than load
sharing. This is the preferred configuration for high availability.

Otherwise, you can look at your switches and either do port trunking or
spanning if your NIC supports it. (Don't do this, just know that its an
option).
 
Ryan Hanisco said:
You want to configure the dual NICs in failover mode rather than load
sharing. This is the preferred configuration for high availability.

Otherwise, you can look at your switches and either do port trunking or
spanning if your NIC supports it. (Don't do this, just know that its an
option).
Click to expand...
Ryan,
Thanks for the response.What I was getting at though,is that these NICs
have been working fine for 2 years and in one day,they both flaked.That
seems very odd to me.I can't
think of anything that would kill 2 NICs like that.Both with the same
symptoms.They could be pinged but the responses
were erratic.Replacing with new single nics fixed the
problem.Serendipity?Maybe.Driver corruption,on both servers,unlikely.We were
having power problems that day
but I have BIG APC backups,and none of the 80 or so boxes on the network had
any NIC failures,A
 
I experienced a similiar issue only to find it was the backbone blad
going bad not the nics


-
wandere
 
Back
Top