Alert: RAD thread
T. Burns
the online book. I'm in single user mode, do mkfs, mount the disk,
start an xfsdump|xfsrestore. I've done this before without a problem.
Today during the xfsdump I get a million error messges scrolling off
the screen that say
Alert: RAD thread held off for more than 32 ms.
I am in single user mode, should there even be any threads? Any idea
what this error means? I tried this on two different octanes in case
it was a hardware problem, no luck.
Dave
Michael Raymond
hardware problem. The RAD kernel thread is part of IRIX's audio driver
and needs to
run every ms. The warning is there to aid in tracking down anything
that might be holding it off. Hopefully you can ignore it.
OOC what version of IRIX was this?
Michael
--
Michael A. Raymond
Core OS Scheduling Group Real-Time Lead
T. Burns
> > I tried this on two different octanes in
case it
> > was a hardware problem, no luck.
> >
> > Dave
>
>Hi Dave,
>
>By no luck, do you mean you got the errors on the other machines or
you
>didn't?
Both machines acted the same, generated the alert: RAD thread message
ad
infinitum.
>
>You don't say what version of IRIX you are using, but these errors
were a
>bug in 6.5.9 and fixed in 6.5.10.
I'm running 6.5.15. Maybe I did something wrong? I was just following
the steps in the online book under the topic "Creating a new system
disk
by cloning". I assumed the two disks were partitioned the same,
because
they both had a new OS installation on them when they were delivered
to
me. So I did
mkfs /dev/dsk/dks0d2s0
mkdir /clone
mount /dev/dsk/dks0d2s0 /clone
cd /clone
xfsdump -l 0 - / | xfsrestore - .
and the errors start showing up at that point.
Do you know of something other than the 6.5.9 bug that would cause
this
error? If the partitioning or volume header were not identical would
that do it? There's some homework I should have done - I wonder if the
SGI site has something that explains obscure error messages.
Thanks again,
Dave