Hi all,
I am currently running an evaluation version of ESXi 5.1 (newest version).
The server hardware is all ESXi certified components.
I have set up Ubuntu 12.04.02 LTS with Kernel 3.5
It seemed to work fine first, but fron the moment on where there's heavy disk I/O strange things happen.
The Kernel is reporting things like:
| [ | 7.102101] Buffer I/O error on device sda3, logical block 32309762 |
| [ | 7.102198] lost page write due to I/O error on sda3 |
| [ | 7.102305] JBD: Error -5 detected when updating journal superblock for sda3. |
| [ | 7.102497] sd 2:0:0:0: [sda] Device not ready |
| [ | 7.102580] sd 2:0:0:0: [sda] |
| [ | 7.102648] Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE |
| [ | 7.102737] sd 2:0:0:0: [sda] |
| [ | 7.102805] Sense Key : Not Ready [current] |
| [ | 7.102932] sd 2:0:0:0: [sda] |
| [ | 7.103000] Add. Sense: No additional sense information |
| [ | 7.103122] sd 2:0:0:0: [sda] CDB: |
| [ | 7.103193] Write(10): 2a 00 10 88 18 18 00 00 30 00 |
| [ | 7.103597] end_request: I/O error, dev sda, sector 277354520 |
| [ | 7.103706] Aborting journal on device sda3. |
| [ | 7.103824] sd 2:0:0:0: [sda] Device not ready |
| [ | 7.103905] sd 2:0:0:0: [sda] |
| [ | 7.103973] Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE |
| [ | 7.104074] sd 2:0:0:0: [sda] |
| [ | 7.104142] Sense Key : Not Ready [current] |
| [ | 7.104273] sd 2:0:0:0: [sda] |
| [ | 7.104349] Add. Sense: No additional sense information |
| [ | 7.104462] sd 2:0:0:0: [sda] CDB: |
| [ | 7.104532] Write(10): 2a 00 10 88 18 10 00 00 08 00 |
| [ | 7.104910] end_request: I/O error, dev sda, sector 277354512 |
| [ | 7.105002] Buffer I/O error on device sda3, logical block 32309762 |
| [ | 7.105108] lost page write due to I/O error on sda3 |
| [ | 7.105205] JBD: Error -5 detected when updating journal superblock for sda3. |
| [ | 7.105366] journal commit I/O error |
| [ | 7.105449] EXT3-fs (sda3): recovery complete |
| [ | 7.105529] EXT3-fs (sda3): mounted filesystem with ordered data mode |
The disk then goes to r/o mode, leading of course to further problems.
The system was initially installed on a VMWare Workstation 9 and then moved to the ESXi server.
It is using the LSI Logic Parallel SCSI controller.
I have searched the web and found almost nothing, except some people back six years ago having had trouble with VMWare Server and a SCSI controller other than LSI Logic Parallel, here http://communities.vmware.com/thread/105970?start=0&tstart=0
The hardware seems fine.
Any help is greatly appreciated.
Attached please find a full dump of the kernel ringbuffer taken via dmesg.
Best,
Tobias