Sun Microsystems, Inc.  Sun System Handbook - ISO 4.1 October 2012 Internal/Partner Edition
   Home | Current Systems | Former STK Products | EOL Systems | Components | General Info | Search | Feedback

Asset ID: 1-72-1353455.1
Update Date:2012-05-14
Keywords:

Solution Type  Problem Resolution Sure

Solution  1353455.1 :   SuSE/Novell SLES10 high average load issue on systems with ILOM hardware (x4x00)  


Related Items
  • Sun 3rd Party Software Product
  •  
  • Sun Fire X4540 Server
  •  
  • Sun Fire X4200 Server
  •  
  • Sun Fire X4200 M2 Server
  •  
  • Sun Fire X4600 M2 Server
  •  
  • Sun Fire X4100 M2 Server
  •  
  • Sun Fire X4100 Server
  •  
  • Sun Fire X4600 Server
  •  
  • Sun Fire X4640 Server
  •  
  • Sun Fire X4500 Server
  •  
Related Categories
  • PLA-Support>Sun Systems>Sun_Other>Sun 3rd Party>SN-OTH: 3PT
  •  




Applies to:

Sun Fire X4100 M2 Server - Version Not Applicable and later
Sun Fire X4100 Server - Version Not Applicable and later
Sun Fire X4200 M2 Server - Version Not Applicable and later
Sun Fire X4200 Server - Version Not Applicable and later
Sun Fire X4500 Server - Version Not Applicable and later
Information in this document applies to any platform.

Symptoms

Customers complaining about a unusual high system load average of about 0.5 to 2.0, which basically means the amount of processes started per second, when running SuSE Enterprise Linux 10 (SLES10). You can see the system load with theuptimeortopcommands.

Changes

This is only happen and true for SLES10 as SLES9 do not have the HAL daemon included.

Cause

This high load is caused by the HAL (Hardware Abstract Layer) daemon hald which polls the devices frequently for media or status changes. The issue is happen when try to polling the virtual USB floppy device provided by the ILOM. There is no mechanism to get the status of the floppy drive, so the hald tries to read from that to see if there is an error (no floppy disk inserted) or not (floppy disk ready). The polling is happen every 2 seconds which cause an average system load from at least 0.5. This behavior is different to a normal floppy drive which will not be polled by the hald.

Solution

The solution is to prevent the haldfrom polling the virtual USB floppy. This can be done by creating the XML configuration file/etc/hal/fdi/policy/10-dont-poll-usb-floppy.fdiwith the following content within the hald configuration:

<?xml version="1.0" encoding="UTF-8"?> <!-- -*- SGML -*- --> <!-- disable polling of USB attached floppies --> <deviceinfo version="0.2"> <device> <match key="@block.storage_device:storage.drive_type" string="floppy"> <match key="@block.storage_device:storage.bus" string="usb"> <merge key="storage.media_check_enabled" type="bool">false</merge> <remove key="info.addons" type="strlist">hald-addon-storage</remove> </match> </match> </device> </deviceinfo>

After restarting the hald(e.g. with the commandrchal restart) the load average go down to normal.


Attachments
This solution has no attachment
  Copyright © 2012 Sun Microsystems, Inc.  All rights reserved.
 Feedback