Isilon OneFS: Node compatibility class create fails when not all drives are HEALTHY

           

   Article Number:     504582                                   Article Version: 3     Article Type:    Break Fix 
   

 


Product:

 

Isilon OneFS 8.1,Isilon OneFS 8.0,Isilon OneFS

 

Issue:

 

 

Creating a node compatibility class fails if not all drives are HEALTHY and causes the process isi_smartpools_d to fail to start. That results in the event:   
    Process isi_smartpools_d of service isi_smartpools_d has failed to restart after multiple attempts   
   
    Running 'isi status -p' will contain the following:   
    Diskpool status temporarily unavailable.   
   
    The following error is logged in /var/log/messages:   

2017-09-08T11:30:59-06:00 <1.4> for-isi-b-1 isi_smartpools_d[5415]: Exception: : Traceback (most recent call last):   File "/usr/bin/isi_smartpools_d", line 287, in <module>     main()   File "/usr/bin/isi_smartpools_d", line 80, in main     run_as_daemon()   File "/usr/bin/isi_smartpools_d", line 89, in run_as_daemon     run_uncaught()   File "/usr/bin/isi_smartpools_d", line 118, in run_uncaught     conform_diskpool_db_to_drive_purpose()   File "/usr/bin/isi_smartpools_d", line 163, in conform_diskpool_db_to_drive_purpose     needs_write = dp_cfg.conform_provisioning_to_node_types(fp_cfg)   File "/usr/local/lib/python2.6/site-packages/isi/smartpools/diskpools.py", line 1200, in conform_provisioning_to_node_types   File "/usr/local/lib/python2.6/site-packages/isi/smartpools/diskpools.py", line 1335, in conform_diskpools_to_storage_units   File "/usr/local/lib/python2.6/site-packages/isi/smartpools/diskpools.py", line 1094, in drive_to_storage_unit AssertionError    
   
                                                                

 

 

Cause:

 

 

A missing drive will cause the disk pool database to fail to update as OneFS is unable to allocate that bay to a disk pool.                                                           

 

 

Resolution:

 

 

   

         
  •         Replace any drives in bays in REPLACE status and make sure all bays in the cluster are HEALTHY. Once all drives are HEALTHY the node compatibility class can be created successfully.     
  •      
  •         After creating the node compatibility class make sure the 'Diskpool status temporarily unavailable' message is no longer in the output of:     
  •    
    # isi status -p   
         
  •         Verify storagepool health and compatible nodes are now in the correct pools by running:     
  •    
    # isi storagepool health -v