VMware Cloud Community
SCC
Contributor
Contributor
Jump to solution

VIC loses connection, can still ping server.

My first go around with installing the ESX 3.02 trial did not go so well (see http://www.vmware.com/community/thread.jspa?threadID=98903&tstart=45) so I reinstalled fresh and eliminated that problem and now have this setup.

Dell 2900

Dual Xeon Quad core

24M RAM

Onboard Broadcom NetXtreme II dual Gbit BCM5708 nic.

port 1 is vmnic0

port 2 is vmnic1

Intel Dual Port 82571EB Gbit nic

port 1 is vmnic2

port 2 is vmnic3

VSwitch0 is the service console on vmnic0 with vswif0 172.28.23.252

VSwitch1 is the VMkernel on vmnic1 with 172.28.23.249

VSwitch2 is the VM Network on vmnic2 and vmnic3.

I have installed the demo licenses as stand-alone.

After about 8 or 10 minutes of using VIC on a windows machine on the same physical network as the server, I get a popup that says "Conneciton to the server has been lost, the application will now exit" and clicking OK closes VIC.

Going to the console and entering 'service network restart' shuts down lo, vswif0 with OK message and then loads 'network settings' with OK and brings lo and vswif0 back up with OK messages and I can connect again with VIC. Then 8 or 10 minutes later, it all starts over.

During all of this I can continuously ping the service console from the VIC client. I do notice that when things are well, the respons is <1ms with a TTL of 64. When the problem occurs, the responses initially jump into 1000+ms with a TTL of 128 and then drop and stay in the 20-40ms range with TTL 128. After the service network restart, the response immediately goes down to <1ms.

I have tried both autoconf and fixed duplex (always getting 100M-Full which is right) on the nics and the switch and am now at fixed.

Short of running a cron job every 5 minutes to restart the network service, any other ideas?

Thanks

Don

Reply
0 Kudos
1 Solution

Accepted Solutions
esiebert7625
Immortal
Immortal
Jump to solution

Based on your ping times and symptoms it sounds like some flaky networking. It could be network cables, physical switch ports or the NIC. I'd start elimimating the possible causes. I'd start with switching the NIC's, try using vmmic2 on your Service Console switch and vmnic0 on your VM Network. If that fixes it it might be a cable/switch port problem.

View solution in original post

Reply
0 Kudos
3 Replies
esiebert7625
Immortal
Immortal
Jump to solution

Based on your ping times and symptoms it sounds like some flaky networking. It could be network cables, physical switch ports or the NIC. I'd start elimimating the possible causes. I'd start with switching the NIC's, try using vmmic2 on your Service Console switch and vmnic0 on your VM Network. If that fixes it it might be a cable/switch port problem.

Reply
0 Kudos
Jae_Ellers
Virtuoso
Virtuoso
Jump to solution

I'm not familiar with Dell's but you want to make sure the Broadcom nic firmware is at the latest. I've had the same issue with HP DL380G4 systems before. I even still have the script around where I unloaded all the network services, restarted the network and reloaded all network services via a script that pinged another host to test connectivity. Ran for months until a firmware update came out.

-=-=-=-=-=-=-=-=-=-=-=-=-=-=- http://blog.mr-vm.com http://www.vmprofessional.com -=-=-=-=-=-=-=-=-=-=-=-=-=-=-
SCC
Contributor
Contributor
Jump to solution

Thanks esiebert7625, you set me on the right track with physical problems. This was the problem.

Don

Reply
0 Kudos