nVidia SATA + 3.5 = Fail

tierstentiersten Member Posts: 4,505
Ugh. I've got a few HP ML115s running as a ESX cluster at home for testing and training purposes. I know its an unsupported configuration by VMware. Its running 3.5u2 and just is unreliable.

I've been having some weird issues with the crappy nVidia SATA controller inside. It is supported by ESX and will run mostly okay with the sata_nv driver apart from randomly it'll just die. The driver will start complaining that the controller had to be taken offline and then basically that node will fall over dead because its lost its HD. Over the space of a couple of days, it'll have this issue. At other times it'll be fine.

Its not the HD either because all the nodes do it. I'd understand if it was a single node that kept dying because of some hardware fault but having that many die is a bit unlikely.

Basically, this post is a gripe about nVidia and VMware :)

scsi: device set offline - command error recovery failed: host 3 channel 0 id 0 lun 0
SCSI disk error : host 3 channel 0 id 0 lun 0 return code = 6000000
I/O error: dev 08:05, sector 43864
I/O error: dev 08:05, sector 43872
I/O error: dev 08:05, sector 44112
I/O error: dev 08:05, sector 44360
I/O error: dev 08:02, sector 3960672
I/O error: dev 08:05, sector 11392
I/O error: dev 08:02, sector 6029328
I/O error: dev 08:02, sector 6029424
I/O error: dev 08:02, sector 6815792
I/O error: dev 08:02, sector 7602224
I/O error: dev 08:02, sector 9699904
I/O error: dev 08:02, sector 9699968
I/O error: dev 08:02, sector 65856
I/O error: dev 08:02, sector 786512
I/O error: dev 08:02, sector 1310840
I/O error: dev 08:02, sector 1310896
I/O error: dev 08:02, sector 2097240
I/O error: dev 08:02, sector 2097272
I/O error: dev 08:02, sector 2097288
I/O error: dev 08:02, sector 2097320
I/O error: dev 08:02, sector 3932176
I/O error: dev 08:02, sector 3932232
I/O error: dev 08:02, sector 3932264
I/O error: dev 08:02, sector 3961640

etc...

Comments

Sign In or Register to comment.