vCenter

Hi All

Experiencing the above issue above on all hosts within a specific cluster on our kit.

Nothing else logged apart from a lot of FT errors (not pertinent) and the following:

"Remote server closed connection after 0 response bytes read" in UM logs

After trying most of the logged fixes, stumbled across this gem:

SELECT ID, DESCRIPTION,ANNOTATION, CONFIG,

• FROM VPX_VM WHERE (CONFIG IS NOT NULL) AND (CONFIG LIKE '%VirtualLsiLogicController%') AND

(ID NOT

IN (SELECT VM_ID FROM VPX_VIRTUAL_SCSICONTROLLER))

Revelead a VM (look in the config column for the FQDN or machine name) with a duff configuration. On selecting it, the vCentre UI hung on loading storage.

Luckily, the machine in question was a test machine so I was able to resolve the issue simply by deleting the machine. There is obviously a more graceful way to do this and clear the rogue entry from the DB (Storage VMotion springs to mind)

Bizarrely this issue was also causing all sorts of HTTP chunking and zero byte responses in the vws.log (VMware VirtualCenter Management Webservices) which was in turn preventing vCentre Service Status and Hardware Staus tabs to fail with:

An error occurred, please try again in another vSphere session.'

We were also getting timeouts from SRM as it synced the storage arrays.

Basically, a single 'stuck' machine was causing all sorts of pain for us as part of our ESX 4.0U1 upgrade and took a while to resolve. I thought I should share it in case you were unlucky enough to experience the same issue!!

Wow, good catch! Tks for sharing this. I've seen corrupt .vmx's do some odd things to host/VC, but this exceeds those experiences.