From: blee@tim.dfrf.nasa.gov (Bryson Lee) Subject: HD Error Question with IDE Drives Date: Fri, 7 May 1993 15:25:41 GMT
If this question has been addressed previously, forgive me...
I'm running an SLS distribution of kernel level 0.99p6 on a 386DX-20, 8MB RAM,
no-name 16-bit IDE Hard/floppy controller with 2 HD:
Master: Seagate ST3144A
Slave: Western Digital 40MB, don't know the model number offhand
On the Seagate Drive, 3 partitions --
/dev/hda1 = 40MB DOS
/dev/hda2 = 65MB Linux/Minix mounted as /
/dev/hda3 = 16MB Swap
On the WD Drive, 1 partition --
/dev/hdb1 = 40MB Linux/Minix mounted as /home
Now for the question:
If I run two disk-intensive operations simultaneously, for example a compile of
some software package, together with a povray raytrace, I almost always see at
some point the following messages echoed to one of my login screens:
HD Error: Read_intr: Status 0x59
HD Error: Read_intr: Status 0x40
Nothing bad seems to happen, even if the messages occur more than once. This
situation usually occurrs when both disk operations access the same physical
drive, most often the WD slave. (That's where I do all my compiling,
user oper-
ations, etc).
Are these messages simply a result of the kernel wanting to do things faster
than my hardware is really able, or are they indicative of some more serious
problem with my kernel/hardware/filesystems? Is there a better (ie faster and
more reliable) alternative filesystem to the Minix FS?
Any enlightenment by mail or followup post would be greatly appreciated!
Bryson Lee
blee@tim.dfrf.nasa.gov