Hello,
As usual, I'll start by saying that it worked before ![]()
The problem occured Monday morning.
The file server in 2012 randomly loses network connection.
This can happen after 2 hours as after 30 minutes.
The file server have also the role of print server.
The virtual server with 8GB of RAM and 3VCPU on ESXi 5.1.
All other servers on the same switch does not have this behavior.
And the cut did not last long between 1 and 3 minutes usually.
When I look in the logs I have absolutely no warning or error and no event tonight warning last happened yesterday and I've been full of problems today deco.
For Info I IPV6 activated as well as domain controllers (2012 + forest and domain).
During this break the net and DNS resolution is ok on computer.
I also tested by removing the virtual network adapter e1000 by replacing a vmxnet3.
Following a VMware and Microsoft KB I also enabled RSS that was not the default.
I also reinstalled the VMware Tools.
I saw on the net as possible resolutions by turning on the server "side received scalling" network or "Checksum Offload" card but I'd rather have an opinion before acting.
TrendMicro Antivirus, Filezilla Server and Backup Exec Agent 2012 are installed on this server.
No other software.
I checked everywhere but then I do not understand at all where it can come.
Thank you in advance for your back because then I stuck serious.
Hello,
Thank's for your reply.
Finally I've not updated it's a network route in linux server which cause an IP usurpation of the IP of FileServer.
(It's a little bit complicated and non logical but when I remove the false route it's working better).
I see on net than VMXNET3 are better support of Aggregation and more...
- how many vNics are connected to your VM network, can you assign only one vNic and observe the problem (do you have any load balancer configured on vNic)
- Did you observe any packet drops or similar problem any other guest OS connected to same VM network.
Currently I have a special vswitch just for this server.
With two vmnic.
This two vmnic are connected to a cisco 3750X stack with port-channel enabled (no lacp).
I have some other server like both exchange, domain controler in the same network with the same vswitch configuration without problem.
I have vmnic unused and I'll try to test with only one vmnic.
Not better.
Each time the latency or cut is very short 30seconds or 1 minute and it's very hard to found this.
Thank's.
I've had a similiar issue with Windows Server 2012 with permanent loss of connection of second adaptor about once a day.
Reason was also "e1000" - which is only available if VM ist set to Server 2008 (VM->Preferences->Options-> Guest OS). To maybe also you have that old setting.
I've figured this out because other Server 2012 didn't hat that problem and this was the only difference (because we installed new Windows on the old VM).
Hello,
Thank's for your reply.
Finally I've not updated it's a network route in linux server which cause an IP usurpation of the IP of FileServer.
(It's a little bit complicated and non logical but when I remove the false route it's working better).
I see on net than VMXNET3 are better support of Aggregation and more...
