Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 28 Oct 1997 06:41:16 -0500 (EST)
From:      Thomas David Rivers <rivers@dignus.com>
To:        perlsta@cs.sunyit.edu, rivers@dignus.com
Cc:        freebsd-hackers@freefall.FreeBSD.org
Subject:   Re: sio silo overflows on a P75 @ 38400 baud?
Message-ID:  <199710281141.GAA06368@lakes.dignus.com>

next in thread | raw e-mail | index | archive | help
> 
> if you've only seen three a day or so, it makes total sense that they
> might be occuring during the nightly/weekly/monthly scripts that seriously
> thrash your system.

 Ummm.... perhaps I mislead... 

 I've wiped a machine and I am reloading it's meaningful stuff over
a SL/IP connection... so, I'm just seeing the ones I'm seeing during
this operation - not over an extended period of time.

> 
> again it's easy to has 3 (or more) situtations, in your situation (not a
> lot of resources) where the system get's overloaded just for a few seconds
> or less causing the overflow...

 This could be the case - but since I'm not doing anything else (although
it is in multiuser) besides flipping syscons virtual screens (not even
typing) I don't believe some other unrelated task has popped in and
is consuming resources.

> 
> it's not more irritating, it's unavoidable. :)
> just think how many interupts there might be in second occuring on your
> serial port, just a little overactivity on the disk..... and wha-la.

 Yes, I can certainly see that - I'm just not sure it's my situation,
it could be, but I'm not aware of the disk over-activity if it's occurring.

	- Thanks -
	- Dave Rivers -

> 
> > > you are quite possbily over loading your system with a combination of
> > > writes and swapping (8megs of ram? ewwww)
> > 
> >  Hmm... good point.
> > 
> > > 
> > > if disk activity is constant it's quite possible to reach a load of 9.0+
> > > i did while doing a buildworld and making my kernel -j8... it was at like
> > > 9.6+ at times... pretty cool as X kept freezing for several seconds at a
> > > time...
> > 
> >  But - the disk activity isn't constant; it's no where near that
> > (being a little twiddle about every 13 seconds). In fact, uptime
> > shows my load average as
> >   7:02AM  up  8:52, 2 users, load averages: 0.00, 0.00, 0.00
> > 
> >  So, I don't believe that's the culprit.
> > 
> >  In the span of this 8:52 hours, though, I've only seen 3 silo
> > overflows...  So, the issue appears to be intermittent... (which
> > makes it more aggravating... :-) )
> > 
> > 	- Dave Rivers -
> > 
> > > 
> > > On Mon, 27 Oct 1997, Thomas David Rivers wrote:
> > > 
> > > > 
> > > > I was just wondering - should it be possible, at 38400 baud,
> > > > in multi-user mode, but nothing else really going on; to get
> > > > silo overflows on a P75 with 16550 (clone?) UARTs?
> > > > 
> > > > I'm doing a SL/IP connection and sending the output of
> > > > dd'ing a tape back to the P75 system for un-tarring.  The
> > > > sending system is a P200 (running FreeBSD 2.2-970510.)
> > > > 
> > > > I'm getting these silo overflows with 2.2.5.
> > > > 
> > > > I'm hoping someone can whip out some figures on the
> > > > interrupt latency to suggest that a P75 should be able
> > > > to deal with receiving 38400...
> > > > 
> > > > This could, of course, be an artifact of some device
> > > > holding the bus too long.  The P75 machine is a laptop
> > > > with a IDE drive (to which I'm writting) and 8 meg of memory;
> > > > again, running 2.2.5-RELEASE.
> > > > 
> > > > 	 - Thanks -
> > > > 	- Dave Rivers -
> > > > 
> > > 
> > > 
> > 
> 
> 



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199710281141.GAA06368>