VMware Cloud Community
tdubb123
Expert
Expert

Migration to host x.x.x.x failed with error Already disconnected (195887150).

I am getting this timeout error in vcenter when doing a vmotion of a vm. any idea what this is?

this is on esxi 5.1 914609

Migration to host x.x.x.x failed with error Already disconnected (195887150).
vMotion migration [171704789:1358656707894315] failed writing stream completion: Already disconnected
vMotion migration [171704789:1358656707894315] failed to flush stream buffer: Already disconnected
vMotion migration [171704789:1358656707894315] socket connected returned: Already disconnected
Reply
0 Kudos
25 Replies
tsanches
Contributor
Contributor

the vMotion vmk and the nfs vmk are in separated vlan's, but are in the same standard vswitch which have 2 uplinks. The uplinks are from 2 separated Qlogic 8152 CNA's.

I already check the driver and firmware version and it is ok. both are in the vsphere 5.1 hcl.

Reply
0 Kudos
spascoe64
Contributor
Contributor

YES!!!!

We were able to get this problems solved by the following:

On the R710 host with the 8152 CNA card do the following:

1.  Disable the MSIX capability

    esxcli system module parameters set -m qla2xxx --parameter-string "ql2xenablemsix=0"

https://support.qlogic.com/app/answers/detail/a_id/1877/related/1/session/L2F2LzEvdGltZS8xMzY2ODQ5NT...

2. Change the network packet pool size

    In vCenter, navigate to the affected Host. Choose Configuration, Choose Advanced Settings under System. Find VMkernel.Boot.netPktPoolMaxSize and change it from 656 to 800.

http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=204287...

3.  Finally, the Most critical change is to double the netPktHeapMaxSize.

   In vCenter, navigate to the affected Host. Choose Configuration, Choose Advanced Settings under System, Change the value of VMkernel.Boot.

netPktHeapMaxSize (under the VMkernel section of Advanced Settings) from 64 to 128
This instruction came from our open ticket with tech support.  We were able to validate that this resolved the problem on our affected hosts.
Scott
Reply
0 Kudos
girish7695
Contributor
Contributor

Try this Simple thing. Reconfigure HA and DRS for cluster. Most of the issues will solve.

Thx

Girish

Reply
0 Kudos
tsanches
Contributor
Contributor

Hey scott, thank you very much!!

this solution worked perfectly. vmotion operations are now flying, and without halting.

Reply
0 Kudos
RonakS
Contributor
Contributor

Hi,

I had faced same issue. As my troubleshooting I came to know this issue is related with HA. I tried to reconfigure the HA on host but didn't worked.

To resolved I removed the HA from host. Rebooted the host & then configured HA on the host.

This happens if HA is unable to updated the record of VM in his protected list.

Regards,

Ronak

Reply
0 Kudos
Bacon_Phat
Contributor
Contributor

I was able to change the vMotion IP, worked like a champ.  Not sure what caused it, but its working fine for me now.  Thought I would post.

Reply
0 Kudos