Re: NTP dumps Linux, film at 11. [Fwd/FYI]

Brandon S. Allbery KF8NH (
Tue, 01 Dec 1998 18:03:07 -0500

In message <199812011740.MAA28826@dcl.MIT.EDU>, "Theodore Y. Ts'o" writes:
| Date: Mon, 30 Nov 1998 17:48:18 -0500
| From: "Brandon S. Allbery KF8NH" <>
| What's involved? We use NTP quite a lot, especially for the Linux
| boxes (RH5 systems tend to drift a *lot* for some strange
| reason...
| Well, one thing which I recently noticed is that if you're using a
| BusToaster SCSI PCMCIA card (great name, isn't it? :-) It's basically

I've got one, although I no longer have a notebook to use it in, and when I
did it ran OS/2.

| So there are at least a few device drivers which are apparently holding
| interrupts off long enough so that we lose a clock interrupt. Not good,

That was my conclusion back in April when I sat down and tried to diagnose
it. There was a little bit of evidence that the console driver was
involved, but not much... and I completely failed to find a common thread
between my own system at home and all the various systems in CMU ECE, aside
from the use of Red Hat 5.x.

Whatever it is, it's been consistent across 2.0.33 - 2.0.36 and it *didn't*
happen on my older system (486DX2/66) with 2.0.33 kernel --- but I've only
observed it on Red Hat 5 and to date I haven't had an RH5 install on the
older system (it ran RH4 for over a year without showing any such symptom,
but the same applies to some older ECE systems including some with hardware
identical to the ones that have severe clock drift).

The other thing about it is that it's not just lost interrupts: the systems
are as apt to *gain* time, instead of losing it, again enough to make
Kerberos upset (I've seen them gain 10 minutes over 5 minutes of wall time).
Again, I haven't been able to correlate gain vs. loss (or amount thereof)
to any particular program or device/driver.

brandon s. allbery	[os/2][linux][solaris][japh]
system administrator	     [WAY too many hats]
carnegie mellon / electrical and computer engineering			 KF8NH
			  Kiss my bits, Billy-boy.

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to Please read the FAQ at