System information

Troubleshooting
6-4
3. Verify hard disk health status
To know exactly if a hard disk fails or not, using SMART check or DST (Device Self-Test) to test the hard disks in question
is a good choice. It’s also advised to check the number of bad blocks and warning events reported by the RAID controller
(see 2.8.1 Hard disks on page 2-38). Another indicator to the health condition of a hard disk is its I/O response time,
because out-of-specification response time could be caused by abnormal error recovery procedures (see 2.11.1 Hard
disks on page 2-54).
4. Adjust hard disk settings
Tweaking the hard disk-related settings could help to accommodate the exceptional behaviors of hard disks (see 2.8.1
Hard disks on page 2-38). The following are some common workarounds:
• Extend Disk I/O Timeout to accommodate slow disk operation
• Increase Disk I/O Retry Time to try I/O more times before giving up
• Reduce Transfer Speed to mitigate bad signal quality of disks
• Disable I/O Queue to avoid problematic NCQ support of disks
• Disable Disk Standby Mode to avoid problematic sleep support of disks
• Extend Disk Access Delay Time to allow longer time for disk spin-up
5. Check failure of multiple hard disks
If multiple hard disks are taken offline at a time, it could be system-level problem. For example, if the hard disks in a
JBOD system are offline unexpected, poor cabling in SAS or FC expansion chain would also lead to unexpected hard
disk offline. In addition, poor heat ventilation, unstable power supply, or hardware quality issues could also lead to offline
of multiple hard disks. In case there are multiple failed hard disks causing corrupted logical disks, you may try to use
array recovery utility to the RAID configurations (see 2.7.13 Array recovery on page 2-35).
6. Ensure proper power on/off sequence of RAID and JBOD
You have to make sure the expansion chassis has been ready before powering on the RAID system, such that the RAID
system can properly recognize the expansion chassis and the hard disks. And shut down the RAID system first and then
its expansion chassis, such that the RAID systems won’t see the lost of expansion chassis as a failure event.
7. AAMUX boards are required for attaching SATA drives to SAS systems
An AAMUX board is required SATA hard disks installed in SAS JBOD systems and for redundant-controller systems.
AAMUX provides dual paths for controllers to access single-ported SATA drives. Check your RAID system hardware
manual for more details, and contact your RAID system supplier to get the AAMUX boards.
8. Regular maintenance helps to avoid disaster
To avoid performance degradation or data loss caused by failed hard disks, it’s advised to enable periodical SMART
checking, automatic disk cloning, and periodical disk scrubbing. Please also set up spare disks for such that disk
rebuilding can be started right away when any hard disks fail.
6.5 User Interfaces
1. The Web GUI cannot be connected
You can use the Web GUI and other network-related management features (including the host-side management
software) only when the management network interface is configured properly. Please check if the LED indicators of the
Ethernet port are lit up properly (refer to the hardware manual for details), and the IP address has to be set to enable the
network interface. You can manually set the IP address by the LCD or local CLI, or use a DHCP server on the network.
2. Some user operations are prohibited
Each user operation can be executed assuming some sort of conditions, please check Chapter 2 to find out the
restrictions of your operations. The web GUI will also offer information and suggestions when a user operation cannot be
done. A few common guidelines and examples are as below:
• To avoid mistakenly destroying data, a disk group cannot be deleted when it contain logical disks.
• To avoid overloading, disk scrubbing is not allowed for a degraded disk group, or an initializing disk group.
• To avoid confusing operations, some settings cannot be modified when a related state happens. For example,
rebuilding-related settings cannot be modified when there is rebuilding disk.
3. The web GUI does not show updated information
The web browser displays the cached the web pages such that you cannot see the updated web pages. For example, the
percentage of firmware upload progress isn’t updated. Please go to the setting page of your web browser to delete the
temporary files, cookies, and the history.