CAM status: SCSI Status Error

B

balanga

I have a VPS running FreeBSD 10.1-RELEASE-p41 on amd64

and see numerous instances of the following errors when running dmesg

Code: 
(da0:mpt0:0:0:0): CAM status: SCSI Status Error
(da0:mpt0:0:0:0): SCSI status: Busy
(da0:mpt0:0:0:0): Retrying command
(da0:mpt0:0:0:0): WRITE(10). CDB: 2a 00 01 fc b9 22 00 00 40 00

Can anyone tell me what is happening?
 
I'm doubtful this is a VPS, they generally don't emulate an LSI Fusion Fibre Channel card. Which VPS provider is this? Or do you have some technical details of their infrastructure (what virtualization technology is used)?
 
balanga said:
I have a VPS running FreeBSD 10.1-RELEASE-p41 on amd64

and see numerous instances of the following errors when running dmesg

Code: 
(da0:mpt0:0:0:0): CAM status: SCSI Status Error
(da0:mpt0:0:0:0): SCSI status: Busy
(da0:mpt0:0:0:0): Retrying command
(da0:mpt0:0:0:0): WRITE(10). CDB: 2a 00 01 fc b9 22 00 00 40 00

Can anyone tell me what is happening?
Click to expand...

Any takers?
 
Make sure VMWare is up to date. And I'd also update the machine to 10.3. FreeBSD 10.1 will be end of life at the end of this year.
 
Not sure how to tell....

vCenter - vSphere Client --->Help-->About VMWare vSphere shows VMWare vCenter Server Version 6.0.0

Hope this helps
 
SirDice said:
Make sure VMWare is up to date. And I'd also update the machine to 10.3. FreeBSD 10.1 will be end of life at the end of this year.
Click to expand...

I'm a bit nervous about doing something like this on a remote VPS. Can you provide a link which explains how to do this?

I guess I should ask the provider to update their FreeBSD offering.
 
balanga said:
Not sure how to tell....

vCenter - vSphere Client --->Help-->About VMWare vSphere shows VMWare vCenter Server Version 6.0.0

Hope this helps
Click to expand...
Yes, I think it is the same version with the one I am having the same issues. I also get a lot of : em0: Watchdog timeout -- resetting

I believe that it is quite safe for you to upgrade. I am running 10.3-RELEASE. If you have the option to take a snapshot before it would be good to use it though.
 
gkontos said:
Yes, I think it is the same version with the one I am having the same issues. I also get a lot of : em0: Watchdog timeout -- resetting

I believe that it is quite safe for you to upgrade. I am running 10.3-RELEASE. If you have the option to take a snapshot before it would be good to use it though.
Click to expand...


I believe I saw the: em0: Watchdog timeout -- resetting message recently before doing a reboot last week.

As for taking a snapshot, how would I go about doing one?
 
I am not sure how. Have you searched for that option in vSphere Client ?
When I perform maintenance on that VM, I usually call the DC and they do it for me.
 
I have had the error

"(da0:mpt0:0:0:0): CAM status: SCSI Status Error
(da0:mpt0:0:0:0): SCSI status: Busy
(da0:mpt0:0:0:0): Retrying command
(da0:mpt0:0:0:0): WRITE(10). CDB: 2a 00 01 fc b9 22 00 00 40 00"

in my FreeBSD daily logs. The issue was on a HP DL380 G7 with vmware esxi, the cause turned out to be the cache controller battery which had failed resulting in disk read times upto 15 seconds. As soon as I replaced the cache controller battery the disk performance returned to normal and the error messages stopped. If the cache controller battery has a solid amber light on the 3rd led suggesting it is faulty. This Steady glow means— There is a short circuit across the battery terminals or within the battery
pack. BBWC features are disabled until the battery pack is replaced.

The only way I could check the battery was to take the cover off the G7 and inspect the LEDS. There was no sign of a problem in iLo


I hope this helps someone else save a bucket load of time.
 
That means that VMware is providing a crappy (or great?) SCSI emulation: They turn a successful but slow SCSI operation (it's perfectly valid for a SCSI Read(10) or Write(10) command to take 15 seconds) into an error. That's either idiocy or genius, depending how you look at it. Personally, I haven't found a SCSI emulation yet that I didn't hate, because they always do such a crappy job of pretending to be SCSI. It seems that the only really good SCSI implementations are real native SCSI devices (Seagate, Hitachi and the like). In particular, Linux is chock full of SCSI emulations (for things like SATA and NVMe) that look like they were written by drunk undergraduates.
 
You must log in or register to reply here.
Back
Top