VMware Cloud Community
struzzo
Contributor
Contributor

ESXi Cluster unrechabel

Hallo,

is it possible that one server which has a PSOD can make a networkimpact of all hosts in the HA cluster?

Thanks for feedback.

Greets Ralf Strauss

0 Kudos
3 Replies
ezzeldin72
VMware Employee
VMware Employee

it depends on your perspective. general speaking No, but you have to consider that, the VMs running on the dead host will be restarted on the other hosts in the cluster and this may show some impact.

Ezzeldin Hussein | MBA| VCAP-DCA/DCD | VCI Level II | VCP-DCV/DT/CMA/NX | VCA/VSP/VTSP | vExpert Team Lead, Systems Engineering, NALE | Member of CTO Ambassador Program.  Business Central Tower A, Dubai Internet City, Dubai, POB 500569 Mobile(EG): +20106 5533 950 Mobile(UAE): +971 56 9095 106 Mobile(OM): +968 9066 0533
0 Kudos
idle-jam
Immortal
Immortal

No, when a single host goes down to PSOD the rest of the hosts will be alive to be the host for the failed VMs to get power up.


iDLE-jAM | VCP 2, VCP 3 & VCP 4

If you found this or any other answer useful please consider the use of the Helpful or correct buttons to award points

0 Kudos
dantyrrell
Contributor
Contributor

Hi,

Yes and No! A very rare scenario - but sure a single rouge NIC on a single host can take out an ethernet network. It needs to play very badly though.

What hardware are you using? Can I guess?

http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=102161...

Possibly fixed by:

http://kb.vmware.com/selfserchecksummingvice/microsites/search.do?language=en_US&cmd=displayKC&exter...

Also see:

http://www.wooditwork.com/2010/11/12/new-broadcom-bnx2x-nic-driver-1-60-is-out/

To quote: "Far more serious was a PSOD issue caused by IP checksumming. The workaround was to disable IP checksumming support but the fix didn’t persist across reboots. "

It looks like disabling IP check-summing offload and TX flow control (if not using iSCSI) might mitigate the issue.

Regards,

Dan

0 Kudos