• Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 777
  • Last Modified:

Can ping SAN from switch, but not from host

We have Dell Poweredge Windows Server 2008 R2 servers. They are connected to a Brocade/Foundry switch. Attached to the same switch in the same VLAN/Subnet is a Dell Equallogic SAN array.

Every day at seemingly random times, for periods of a few hours, our servers can no longer ping the SAN which is on the same switch. However if I telnet to the switch I can ping the SAN. One might think the switch is blocking them, but yet the SAN volumes are up and data is being written to them, so I know the switch isn't blocking them. We've tried rebooting the SAN, upgrading to the latest firmware, etc. We haven't replaced all of the cables (which are a couple months old), nor have we changed all 4 ports the cables are plugged into. It just seems very unlikely that all 4 of cables or ports that happen to be for a single member would die. We are working with Dell from the SAN and server perspective, but I'm trying to think of anything on the switch that could be the cause. We cannot really reboot the switch as it serves up many other connections/devices besides this SAN.
0
MrVault
Asked:
MrVault
  • 6
  • 4
3 Solutions
 
SouljaCommented:
I don't think it is anything on the switch, unless your switch is layer 3 and you have some ACL applied. I would check the cables, make sure you are using a high quality cable. I remember at one time replacing every cable to a SAN because the cables were of low quality. Interfaces kept dropping randomly. We put in high quality cables and never had another problem.
0
 
MrVaultAuthor Commented:
how do I tell if spanning tree is getting in the way?

here's what's weird. there are 2 enclosures (members) in the SAN group. they share a single group IP and each of the 4 NICs in each member has an IP. At first the 4 in member1 were not pingable except if I was directly on the switch. I unplugged those 4 cables and put them in  4 different ports on the same switch and then they all came up, but 2 on the other member went down. for the last hour those original 4 that were down are up, and 2 new ones are down. these cables are the same type/brand/order as all the others in our network that have never had this problem in 2 years. the cables from all 8 are about 3 months old.

the SAN vendor is thinking it's a switch problem. What's weird is that if you telnet to the switch it can ping all 8. However if you go to a host on the same VLAN plugged into the same switch, that host cannot ping those 2. However if you telnet to another SAN array plugged in 5 switches away, that CAN ping all 8 ports. So we know the switch isn't completely blocking it and the cables are working.
0
 
SouljaCommented:
I'm not familiar with the Brocade switches, but can you look at the status of the ports in spanning tree and see if they are in blocking state when this occurs?
0
Industry Leaders: We Want Your Opinion!

We value your feedback.

Take our survey and automatically be enter to win anyone of the following:
Yeti Cooler, Amazon eGift Card, and Movie eGift Card!

 
SouljaCommented:
Being that you have muliple ports plugging into the switch from the same host, it could possibly be a spanning tree issue.
0
 
MrVaultAuthor Commented:
they are forwarding always.
0
 
MrVaultAuthor Commented:
the 4 NICs on the array each have their own IP and there is nothing else on the network with the same IP.
0
 
SouljaCommented:
It's not a spanning tree issue if they are remaining in forwarding state.
0
 
MrVaultAuthor Commented:
yeah, that's what we were thinking.

so the other 2 that started having the issue after I updated the others, I've moved those 2 cables into two other free ports on the switch and they are now connecting just fine. we'll have to see if these stay up now. i'll let you know.
0
 
MrVaultAuthor Commented:
we had about 10 more start having the same problem. we replaced the switch (it was an ebay refurb) and so far no issues. bad hardware is likely.
0
 
MrVaultAuthor Commented:
was hardware issue. could have been a firmware issue, but was not able to get update service to latest code. replaced switch.
0

Featured Post

[Webinar] Database Backup and Recovery

Does your company store data on premises, off site, in the cloud, or a combination of these? If you answered “yes”, you need a data backup recovery plan that fits each and every platform. Watch now as as Percona teaches us how to build agile data backup recovery plan.

  • 6
  • 4
Tackle projects and never again get stuck behind a technical roadblock.
Join Now