mkibler@eecs.wsu.edu (Michael K. Kibler) (11/03/90)
We have been receiving the following error on our HP9000 840. usrpt: rmap ovfl [1469,1470) Rebooting the system clears the error. But it would be nice to know what the error is referring to and possible causes. Any ideas??? Thanks in advance, -- ---- Mike ( ~~ Radiosity is more than just heat! ) - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Michael K. Kibler INTERNET: mkibler@eecs.wsu.edu Elect/Compt. Engr. Dept. UUCP : ...uunet!eecs.wsu.edu!mkibler
Kimmo.Suominen@lut.fi (Kimmo Suominen) (11/09/90)
>>>>> On 2 Nov 90 21:40:28 GMT, mkibler@eecs.wsu.edu (Michael K. Kibler) said:
Michael> We have been receiving the following error on our HP9000 840.
Michael> usrpt: rmap ovfl [1469,1470)
We get that too on our HP9000/845. It runs HP-UX 7.0 with ARPA and
NFS. I can't recall getting it in 3.10.
Also, rebooting helps only for a short while.
It would REALLY be nice to know why, though it hasn't bothered me
enought yet to make me call the local support - maybe I'll do that
tomorrow. The mismatched pair of parentheses/brackets is really cute
;-)
--
Kim / Internet: Kimmo.Suominen@lut.fi
"That's what I think." / Bitnet: KIM@FINFILES
frank@hpwrce.HP.COM (Frank Stutzman) (11/10/90)
Michael> usrpt: rmap ovfl [1469,1470) Kim> Kim>We get that too on our HP9000/845. It runs HP-UX 7.0 with ARPA and Kim>NFS. I can't recall getting it in 3.10. Kim> Kim>Also, rebooting helps only for a short while. Rebooting works fine until someone starts up monitor again. Monitor spits out this error (or at least one close to it) when it is run on a system that has had some tunable parameters changed (noteably NPROC>512). This is a known problem with monitor at 7.0, but you cannot count on it to be fixed as monitor is not an supported utility. Its only exists due to the willingings of a engineer in the kernal lab to spend his weekends tweaking it. |=============================================================================| |Frank Stutzman | "What is wanted is not the will | |Hewlett-Packard Western Response Center | to believe, but the will to | |Mtn. View, Ca | find out, which is the exact | |frank@hpwrc.hp.com | opposite." -Bertrand Russell | |=============================================================================|
peke@nesterc.uucp (Pekka Keskiivari) (11/12/90)
In article <KIMMO.SUOMINEN.90Nov8175329@kannel.lut.fi> Kimmo.Suominen@lut.fi (Kimmo Suominen) writes: >>>>>> On 2 Nov 90 21:40:28 GMT, mkibler@eecs.wsu.edu (Michael K. Kibler) said: > >Michael> We have been receiving the following error on our HP9000 840. >Michael> usrpt: rmap ovfl [1469,1470) > >We get that too on our HP9000/845. It runs HP-UX 7.0 with ARPA and >NFS. I can't recall getting it in 3.10. > >Also, rebooting helps only for a short while. > >It would REALLY be nice to know why, though it hasn't bothered me >enought yet to make me call the local support - maybe I'll do that This is probably caused by a bug in the Monitor program (SR/KPR# 4700-870915). It causes the virtual memory to be gradually eaten up... The kernel exhibits this behaviour if NPROC happens to be > 512. As you might suspect, there is a patch available from HP. It replaces pstat.o in /etc/conf/lib/libhp-ux.a. >tomorrow. The mismatched pair of parentheses/brackets is really cute ;-) The numbers might stand for a semi-closed range of memory addresses? -- Pekka Keskiivari, Neste Oy Technology Centre peke@neste.fi
Kimmo.Suominen@lut.fi (Kimmo Suominen) (11/15/90)
>>>>> On Mon, 12 Nov 90 09:37:04 GMT, peke@nesterc.uucp (Pekka Keskiivari) said:
Pekka> This is probably caused by a bug in the Monitor program
Pekka> (SR/KPR# 4700-870915). It causes the virtual memory to be
Pekka> gradually eaten up... The kernel exhibits this behaviour if
Pekka> NPROC happens to be > 512.
Pekka> As you might suspect, there is a patch available from HP. It
Pekka> replaces pstat.o in /etc/conf/lib/libhp-ux.a.
Well, yes. I have been told that the bug is actually in some BSD 4.2
derived code in the kernel. I got the patch from HP (I guess - it
arrived by e-mail) and I yet need to install it. Seems like monitor
is the only program to bring up this bug.
Thanks for the answers I received!
--
Kim / Internet: Kimmo.Suominen@lut.fi
"That's what I think." / Bitnet: KIM@FINFILES