Hello,
I just compiled kernel 2.6.15. I came across an option that I didn't have any knowledge about:
Block Layer-> IO Schedulers-> < > Anticipatory I/O scheduler < > Deadline I/O scheduler < > CFQ I/O scheduler Default I/O scheduler (CFQ)-> ( ) Anticipatory ( ) Deadline ( ) CFQ ( ) No-op
Here's what the help text in menuconfig says about the three options:
Anticipatory:
CONFIG_IOSCHED_AS: The anticipatory I/O scheduler is the default disk scheduler. It is generally a good choice for most environments, but is quite large and complex when compared to the deadline I/O scheduler, it can also be slower in some cases especially some database loads.
Deadline:
CONFIG_IOSCHED_DEADLINE: The deadline I/O scheduler is simple and compact, and is often as good as the anticipatory I/O scheduler, and in some database workloads, better. In the case of a single process performing I/O to a disk at any one time, its behaviour is almost identical to the anticipatory I/O scheduler and so is a good choice.
CFQ:
CONFIG_IOSCHED_CFQ: The CFQ I/O scheduler tries to distribute bandwidth equally among all processes in the system. It should provide a fair working environment, suitable for desktop systems.
The help text wasn't much help for me either, so I went about googling. I found this:
http://www.linuxhowtos.org/System/iosched.htm
This is very good information.
Anyway, now I'm wondering how much effect an I/O Scheduler will have on the drive performance, as compared to hdparm settings.
'hdparm -tT' gives me a read speed of ~55MB/s on my AMD Sempron64 2500+ system. My hard disk isn't even SATA. I don't think it is possible to squeeze out much more. I suppose switching I/O Schedulers on my system won't make much of a difference.
Does anyone have more information about this?
Mrugesh