IBM 86605SU Maintenance Manual - Page 125
Changing the Write Policy, Read Ahead, Enter.
UPC - 087944475767
View all IBM 86605SU manuals
Add to My Manuals
Save this manual to your list of manuals |
Page 125 highlights
F6: When a drive is removed from the server (for example, for maintenance or security), but is returned to a different bay, the ServeRAID controller adjusts the configuration to the drive's new location. F7: A ServeRAID controller, with stored configuration information that does not match that of the configured drives present, imports the configuration information from the drives. Read Ahead: The default setting is On. Normally, the ServeRAID controller transfers data from disk to its local cache in increments equal to the stripe-unit size. This provides excellent overall performance when workloads are steady and sequential. However, if the workload is random or the system I/O requests are smaller than the stripe-unit size, reading ahead to the end of the stripe might degrade performance. When read ahead is set to Off, the ServeRAID controller transfers data from disk to its local cache in increments equal to the system I/O request size, without reading ahead to the end of the stripe. You can change the read-ahead setting without destroying data in a logical drive. 4. Use the Up Arrow (↑), Down Arrow (↓), Right Arrow (→), or Left Arrow (←) key to highlight the parameter that you want to change; then, press Enter. 5. Press Esc to return to the Advanced Functions menu. 6. Back up the disk-array configuration information to diskette. See "Backing Up the Disk-Array Configuration" on page 126 for instructions. (The backup procedure also saves the RAID parameters to diskette.) Changing the Write Policy: When you configure a logical drive, the ServeRAID controller sets the write policy to write-through (WT) mode, where the completion status of a write command is sent after the data is written to the hard disk drive. Under certain workloads, you can improve performance by changing the write policy to write-back (WB) mode, where the completion status is sent after the data is copied to cache memory, but before the data is actually written to the storage device. Although you might gain performance with write-back mode, it creates a greater risk of losing data due to a power failure. If a ServeRAID controller failure occurs while data is in the cache the data is lost. This is because the system gets a completion status message when the data reaches cache memory, but before data is actually written to the hard disk drive. 120 Netfinity Server HMM