2010-07-02 08:24 AM
Riddle me this - there is a FAS2050A system currently running 126.96.36.199. System Manager was reporting the system's health as green. Over the weekend there was a disk failure, the disk was replaced and assigned to the proper controller. A sysconfig -r reports each controller has a spare of the only disk type (300GB SAS) in the system. Looking under Storage > Disks in System Manager reports the same thing. Yet, System Manager now keeps reporting that there are insufficient spare disks on the controller that had the disk failure. The syslog is not logging any such errors as it normally would if the system spares fall below the min spare count. To verify the raid.min_spare_count is currently set at 1 and there is 1 spare disk. Uninstalling and re-installing System Manager doesn't clear the error, removing and re-adding the system proves just as futile. I'm completely confused as to where it's pulling this information from.
Anybody have any ideas? Perhaps it just isn't playing well with the older version of OnTap on this system?
2010-07-02 11:11 AM
Hi Chris -
System Manager uses the aggr-check-spare-low API, which states: Return true if there is no suitable spare disk available for any filesystem (parity or data) disk.
If you have a client and the ONTAP SDK, you can easily use apitest to verify:
apitest toaster root <password> aggr-check-spare-low
If not, go to the storage controller console and do this:
toaster> priv set advance
toaster*> ontapi execute <aggr-check-spare-low>
<?xml version='1.0' encoding='UTF-8' ?>
<!DOCTYPE netapp SYSTEM '/na_admin/netapp_filer.dtd'>
<netapp version='1.1' xmlns='http://www.netapp.com/filer/admin'>
toaster*> priv set admin
to verify that the result is coming from the aggr-check-spare-low API.
It would be useful to know what version of ONTAP the problem is occuring.
I have updated a bug report.
- Rick -
2010-07-18 07:25 PM
Sorry for the delayed response. The version of OnTap per my post is 188.8.131.52. Unfortunately the box is at a customer site so I don't have direct access. I'll see if he'll run the command for me and report the results.