erc@pai.UUCP (Eric Johnson) (12/05/89)
A colleague asked me to post a message requesting more
information on some strange behavior we have noticed
on an HP 370, running X11R2 (actually, HP's implementation of X
that came with the 370) under HP-UX 6.5:
The problem:
Hpterms, if left alone for a period of time, just seem to
"die".
a) The hpterm window disappears.
b) The process listing in "ps -ef" shows the process as <defunct>.
c) The ps -ef listing shows no trace of the process (a short while
after the <defunct> entry is seen).
(Note: Hpterm is an Hp program that acts much like xterm, but
acts as an HP terminal instead of a pseudo-vt100.)
The term "left alone for a period of time" means that the window has
no input (nor output) for a period of about an hour. The window in
question is usually a "console" window (e.g., the hpterm is called
with the -C option to capture console output). This window is
usually 5 or so lines long and exists mainly to display console
messages. (At other times, larger hpterms "die" as well.) We first
noticed this when a demo was set up. We went away, and when we
came back--much to our surprise--we were missing an hpterm window.
This problem does occur also when a knowledgeable user is sitting
at the console (eliminating my first suspicion of accidental
removal of the hpterm).
We are running a stock HP 370, with an HP SCSI disk (300 MB), the HP
colour monitor and 8 MB RAM, under HP-UX 6.5, HP/NS-ARPA Services
and HP X Window System (all stock that we purchased together
with HP-UX 6.5). All equipment in use on this machine is
stock HP equipment, so I don't think any third-party compatibility
issues are involved here.
Does anyone have any information about this? I'm assuming either
the client (hpterm) or the server (the X server) side is deciding
(wrongly) that communication has broken down, and is cutting the
still live connection. Is this a known bug? (Please don't tell
me to read those terrible bulletins from HP--the information
is in a form that makes them essentially worthless--I gave up on
them back when I was still using the HP-1000.) Is there anything
we can do locally to work around this problem? (I would prefer
that the hpterms remain live until terminated.)
We do not have the HP X11R2 sources (as HP made a number of
changes, I assume). We do have X11R3 (which is much better
than R2 in my opinion) from MIT, but would prefer to use an
HP-supported product (I'm hoping that an HP-version of X11R3
will appear sometime soon--with HP-UX 7.0, but since Minnesota
seems to be in the boonies as far as HP is concerned, I suspect it
won't come in time for us to be able to use it on this project).
If you have any information on this, I would certainly
appreciated it. If I receive a number of replies by email,
I will post a summary.
Thanks,
-Eric
--
Eric F. Johnson, Boulware Technologies, Inc.
415 W. Travelers Trail, Burnsville, MN 55337 USA. Phone: +1 612-894-0313.
erc@pai.mn.org - or - bungia!pai!erc
(We have a very dumb mailer, so please send a bang-!-style return address.)fkittred@bbn.com (Fletcher Kittredge) (12/05/89)
I posting this rather than sending it because this is such a frequently asked question... [ Paraprase: My friend gets logged out from an HP-UX session after some amount of idle time.] Let me guess your friend uses csh. If so, this is documented behavior. Check the csh manual page for the section on the variable "autologout". If you don't want to be automatically logged out, then set "autologout" to zero. regards, fletcher Fletcher E. Kittredge fkittred@bbn.com
andrew@hpqtdla.HP.COM (Andrew Mackenzie) (12/05/89)
I have very little knowledge of X, but a little of using HPterm. Is it possible that your hpterm window is running a shell that has 'auti-logout' set. This can be used by any shell running (i.e. on a normal character terminal too) so that if no process is running and no keys are hit for the specified time then the shell 'logs itself out'. This can be set to 60 minutes or whatever, and if you forget to logout when you go home, you can rest assured that it has been done for you. If this happened on an hpterm window the session would finish and the window would close up and disapear. (only an idea !!!)
campbelr@hpclove.HP.COM (Bob Campbell) (12/06/89)
I suspect that someone from Corvallis will know of any X problems and will reply, but had to check this possibility . . . Do you have a value set for the shell variable autologout? You sound like you have probably checked that, but I was reminded just yesterday how much time can go into a simple problem :-( -------------------------------------------------------------------------- Bob Campbell Some times I wish that I could stop you from campbelr@hpda.hp.com talking, when I hear the silly things you say. Hewlett Packard - Elvis Costello
steve@hpcvlx.cv.hp.com (Steve Hiebert) (12/06/89)
You don't mention what shell you are running in the hpterm window. It sounds to me like your shell is timing out. If you are using csh(1), the pre-defined timeout variable is "autologout" and it is initialized to 60 (minutes). After 60 minutes with no input the shell logs out. Set it to 0 (in your ~/.cshrc or interactively) to completely disable the feature. I.e.: set autologout=0 in your ~/.cshrc or at the keyboard. Steve Hiebert Hewlett-Packard Co.
rjn@hpfcso.HP.COM (Bob Niland) (12/07/89)
re: > Hpterms, if left alone for a period of time, just seem to "die".
Depending on what shell you use, check the "autologout" or "TMOUT"
environment variables. These specify the amount of time that must
elapse after last I/O before the window shell logs itself out.
My windows die too, just when I want them to.
Regards, Hewlett-Packard
Bob Niland rjn%hpfcrjn@hplabs.HP.COM 3404 East Harmony Road
UUCP: [hplabs|hpfcse]!hpfcla!rjn Ft Collins CO 80525-9599steve-t@hpfcso.HP.COM (Steve Taylor) (12/07/89)
} Hpterms, if left alone for a period of time, just seem to "die". } a) The hpterm window disappears. When this happens, is the default shell csh ? The description sounds like what happens when the shell which is spawned for the window terminates. Csh has an autologout feature. From csh(1): If the terminal remains idle (no character input) at the shell's top level for a number of minutes greater than the value assigned to _autologout_, you are automatically logged off. } The term "left alone for a period of time" means that the window has } no input (nor output) for a period of about an hour. Continuing from csh(1): The initial value of _autologout_ is 60. If unset or set to 0, _autologout_ is entiredly disabled. I use ksh on my system, which is also standard 6.5, and have not observed this behavior, but I remember this question or a similar one being asked internally some time back. Regards, Steve Taylor NOT A STATEMENT, OFFICIAL OR OTHERWISE, OF THE HEWLETT-PACKARD COMPANY.
rodean@hpfcdc.HP.COM (Bruce Rodean) (12/08/89)
In article <772@pai.UUCP> erc@pai.UUCP (Eric Johnson) writes: > The problem: > > Hpterms, if left alone for a period of time, just seem to > "die". > a) The hpterm window disappears. > b) The process listing in "ps -ef" shows the process as <defunct>. > c) The ps -ef listing shows no trace of the process (a short while > after the <defunct> entry is seen). > The term "left alone for a period of time" means that the window has > no input (nor output) for a period of about an hour. Are you by any chance using csh with autologout set? It sure sounds like it, especially since the window is idle for about an hour. If so, then unset or set to 0 the shell variable autologout. See the section called "Autologout" on the csh manual entry for more details. Bruce Rodean rodean%hpfclg@hplabs.HP.COM This does not represent an official response of Hewlett-Packard Co. It is my guess only.
stroyan@hpfcdq.HP.COM (Mike Stroyan) (12/08/89)
> The problem: > > Hpterms, if left alone for a period of time, just seem to > "die". Your shell is timing out and exiting. You want to disable the shell's autologout timer. For csh, add set autologout=0 to your .login or .cshrc file. For ksh, add TMOUT=0 to your .profile or .kshrc file. Mike Stroyan, stroyan@hpfcla.hp.com
jack@hpindda.HP.COM (Jack Repenning) (12/08/89)
If you problem is not the autologout business, then it's not a problem
that I've ever seen or heard of, and I keep windows open (including
console windows such as you describe) for days and days, in
essentially the same kind of system as you describe. Under the
assumption, then, that we're dealing with an autologout situation,
I'll also point out that the autologout business can also happen to a
user whose shell is /bin/ksh. In this case, it's controlled by the
variable TMOUT. From the on-line man page for ksh:
TMOUT If set to a value greater than zero, the
shell will terminate if a command is not
entered within the prescribed number of
seconds after issuing the PS1 prompt. (Note
that the shell can be compiled with a maximum
bound for this value which cannot be
exceeded.)
-------------------------------------------------------------
Jack Repenning - Information Networks Division,
Hewlett Packard Company
RFC-822:... jack@hptnjar.cup.hp.com
uucp: ... {allegra,decvax,ihnp4,ucbvax} !hplabs!hptnjar!jack
HPDesk: Jack REPENNING /HP6600/UX
USMail: 43LN; 19420 Homestead Ave; Cupertino, CA 95014
Phone: 408/447-3380 HPTelnet: 1-447-3380
-------------------------------------------------------------
By the way, netnews is a very slow and capricious way to get questions
answered. If you have either HP TeamLine or HP ResponseLine software
support services, you can direct questions about supported HP products to
your regional HP Response Center. HP people who respond to netnews
questions are donating our time. We not have the full resources of a
Response Center at our disposal, and the person who knows the answer to any
given question may not be reading news this week, or month, or...rbent@ge.UUCP (Ray Bentz) (12/08/89)
Is it possible that you are using /bin/csh as your shell and have "autologout" set? Ray Bentz GE Aerospace Mail Stop 138-202 Moorestown, NJ 08057