One of my hosts running 5.1 was recently having some issues with slow-performance on VMs, random issues with vMotion, etc. I began to suspect the Broadcom 10GB NIC. I swapped the NIC yesterday. Rebooted and vSwitch all looked good, etc. However, after this hardware change and reboot, the host is now showing "No vmkcore disk partition is available..." and also messages about not being able to log coredumps.
I am not sure if this is pointing to an underlying hardware issue (possible bad motherboard or backplane?) The host boots fine but continues to have issues scanning for iSCSI data stores (takes much longer to scan for storage than my other identical hosts)
This is a Dell R710 with all the latest/greatest BIOS, firmware and also updated ESXi 5.1 patches. I have four other identical hosts in my cluster that do not have these issues.
At this point, I see a couple options:
1) Call VMWare support and troubleshoot why this vmkcore error suddenly appeared. I am mainly interested in this route to see if we can blame the hardware. In that case, I call Dell with my findings.
By the way, all Dell hardware checks and diags came back as Good including a full memtest on the host that also came out good.
2) Remove the host from vCenter and re-install a fresh copy of ESXi and see where we are at from there.
Reaching out to the community to see if others have had random performance issues and how/if you were able to determine hardware vs. software issue.
