Hi, Kevin
I've tried out all this before and now again... but nfydata.0 is increasing.
After reading again how ntserverd/ntsendd works I thought that the problem
is in nfylogu file... perhaps if I delete all nfy* files?
But, how can I test what is happening... I can't start in debug mode!
Perhaps I'm doing something wrong. When I try to start I receive a message
saying there's already a process with this name, so...
I tried also to kill the notifier processes and then restart: ntsendd seems
to start ok but ntserverd doesn't!!!
?
> Björn Calabuig Plein (NET)
> SSDI - Técnico Funcional Remedy
>
> MENTA
> Lepanto, 350
> 08025 Barcelona, España
> Tel. (+34) 93 530 35 89
> mailto:bcalabuig@ctc.es
> http://www.menta.net
-----Mensaje original-----
De: Kevin Murray [mailto:kevin.murray@FUSION.IE]
Enviado el: jueves 20 de diciembre de 2001 11.29
Para: ARSLIST@LISTSERV.VISTAIT.COM
Asunto: Re: High Availability Problem
**
Hi Bjorn,
Hopefully some insight on one or two of your queries.
Why are there 2 notification log files locations ?
Well the reason being is probably to do with the NTSPOOLDIR environment
setting. By default this env is not set and the notification system defaults
to
/usr/spool/remedy [/var/spool/remedy] but you can point it to a different
location using this ENV var. Perhaps in your case location 1
Perhaps which ever user runs remedy [appears root] has this in their profile
or it is part of a script which you use to start remedy in one
instance, perhaps another script on the system doesn't have it which may
have been used to start remedy since the original crash.
I have worked on HP-ServiceGuard before and using the Remedy specific env
var's was a must as some of the system locations in my experience where not
set to fall
over to node 2 etc... I have also seen sites neglect this and the same file
can appear in numerous locations on the system[s]
As for starting/stopping the notifier it should be possible
ps -ef | grep -i ntserverd
returned
ps -ef | grep
This should return the listing for ntserverd & ntsendd
kill -9
This should also stop the ntsendd if not do a similar cmd for that
To restart
in the appropriate shell as the appropriate user who runs remedy [seems to
be root in your case]
./ntserverd &
This should also start ntsendd
I hope this helps. If there is problems if may be library related and
therefore maybe an env var issue.
"Killing the Notifier processes and executing ./ntserverd -h 1 seems
not to
work."
May it isn't maybe it is. There is a specific algorithm for this which you
can find in the K/B Base and it may not match how you believe it works..
The fact that no notifcations are sent you will probably need to start the
notification system in debug to at least give you
an idea why not.
Regards,
Kevin
> -----Original Message-----
> From: Bjorn Calabuig [SMTP:bcalabuig@CTC.ES]
> Sent: 20 December 2001 09:28
> To: ARSLIST@LISTSERV.VISTAIT.COM
> Subject: RV: High Availability Problem
> Importance: High
>
> Dear List!
> Yesterday I send you a mail explaining a HA problem. But I forgot to
> include
> some platform info. See bottom of mail. I attach you the document again.
>
> <>
> I would like to commentate an interesting situation. Is it a normal
> situation to have two nfydata.0 in two different locations?. I discovered
> this situation on my Production server:
>
> 1 /opt/ars/NotifierFields
> -rw------- 1 root root 399359 Dec 20 09:59 nfydata.0
> -rw------- 1 root root 0 Dec 20 09:00 nfylock.0
> -rw------- 1 root root 1552 Dec 20 09:46 nfylogin.0
> -rw------- 1 root root 5502 Dec 14 15:34 nterror.log
>
> 2 /var/spool/remedy
> -rw------- 1 root sys 57101573 Dec 14 15:13 nfydata.0
> -rw------- 1 root sys 0 Nov 12 08:38 nfylock.0
> -rw------- 1 root sys 0 Dec 14 15:08 nfylogin.0
> -rw------- 1 root sys 3348 Dec 14 15:08 nfylogu
> -rw------- 1 root sys 1251 Dec 14 15:15 nterror.log
>
> We had a crash Dec 14, and on location #2, the nfydata.0 don't grow
> anymore.
> Is on location #1 where this file is growing each day. We have to purge it
> for not filling up the file system. But I don't understand why this file
> is
> is two locations.
>
> Anyway, since the crash the Notifier processes are up but no notification
> is
> sent. Perhaps after this crash this processes don't work well. We have
> here
> an 24x7 environment so, stoping and starting will be difficult.
>
> Is there a way to stop and start only the Notifier processes?
> I tryed to execute ntserverd -d 1 for entering in debug mode, but I get an
> error showing that there's already an process with the same name working.
>
> Killing the Notifier processes and executing ./ntserverd -h 1 seems not to
> work.
>
> I would appreciate any help.
> TIA...
>
> Happy Xmas!!!
> ;P
>
> Server Information:
> Version: 4.00.02
> HW: 9000/800
> OS: HP-UX B. 10.20
> DB: Oracle 8.0.5
>
> > Björn Calabuig Plein (NET)
> > SSDI - Técnico Funcional Remedy
> >
> > MENTA
> > Lepanto, 350
> > 08025 Barcelona, España
> > Tel. (+34) 93 530 35 89
> > mailto:bcalabuig@ctc.es
> > http://www.menta.net
> >
> > << File: HighAvailability.doc >>
**********************************************************************
This email and any files transmitted with it are confidential and
intended solely for the use of the individual or entity to whom they
are addressed. If you have received this email in error please notify
the system manager.
This footnote also confirms that this email message has been swept by
MIMEsweeper for the presence of computer viruses.
www.mimesweeper.com
**********************************************************************
________________
_____________
You may also control your subscription options, including UNSUBSCRIBE, at
www.ARSLIST.org
_____________________________
Previous Topic
Index