SCSI access creating lost time

Buddha Buck (bmbuck@acsu.buffalo.edu)
Sat, 06 Mar 1999 20:25:14 -0500

I've been running into a problem recently with lost time: My system
clock has been falling behind drastically -- earlier this week, with an
uptime of less than 2 days, had lost 5100 seconds.

It seems to be tied to SCSI access:

# netdate tick; dd if=/dev/hda of=/dev/null bs=1kk count=100; netdate
tick
tick +0.223 Sat Mar 6 19:59:08.000
100+0 records in
100+0 records out
tick -0.020 Sat Mar 6 20:00:28.000
# netdate tick; dd if=/dev/sda of=/dev/null bs=1kk count=100; netdate
tick
tick +0.365 Sat Mar 6 20:01:29.000
100+0 records in
100+0 records out
tick +52.664 Sat Mar 6 20:02:49.000

tick is a local Stratum-2 timeserver.

I can deal with the sub-second jitter between calls to netdate, but I
can't deal with the loss of 52 seconds in an 80-second interval! xntp3
can't deal with it, either.

I don't think this is a user-space problem...

I dropped back to Linux 2.2.1 to see if it was a new problem in 2.2.2,
but it doesn't seem to be.

My system:
Linux 2.2.1 (Debian distribution, "potato")
AMD K6-2 333
FIC VA-503+ MB
32MB PC100 SDRAM
Adaptec 1520 SCSI controller
/dev/sda: Seagate ST51080N
/dev/scd0: SONY CD-ROM CDU-76S

Is there any other information that could help?

-- 
     Buddha Buck                      bmbuck@acsu.buffalo.edu
"Just as the strength of the Internet is chaos, so the strength of our
liberty depends upon the chaos and cacaphony of the unfettered speech
the First Amendment protects."  -- A.L.A. v. U.S. Dept. of Justice




-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/