Well, if you have a core then I suggest a call to Support as I did before.
As I tried to explain the ovw command fails with the "cannot connect to
database" because the ovwdb daemon is not running. Until he is running and
ovtopmd is running, ovw will fail. The netview command, on the other hand,
will start these daemons, if they can be started, when you run it from the
command line as root.
The database command ovtopofix does require that ovwdb and ovtopmd be running.
These are the daemons which will carry out the command.
So what you are saying is that if you clear your databases and start over, then
you can get the GUI up. But then as soon as try to reboot the box, you get this
problem and have to clear them again? Did you try formatting the nettl logs as
I suggested? There may be a clue in there.
In any case, I suggest a call to Support so you can get personalized help.
James Shanks
Tivoli (NetView for UNIX) L3 Support
fabiana AT telesc.com DOT br on 10/25/99 01:29:59 PM
To: james_shanks AT tivoli DOT com
cc: (bcc: James Shanks/Tivoli Systems)
Subject: Re: ovwdb daemon
The ovw command the Netview starts for few seconds and after close
the window with error: ovw: cannot to connect the database, but if I try to
rebuild the
database the daemons start and the Netview GUI open.
I can't start the daemons after the reboot from command line. I found the
core file for ovwdb in /usr/OV/PD/cores.
Some command about database I can't do because shows cannot to
connect the database.
I know that the ovwdb cache size is 10.000
# df -k
Filesystem kbytes used avail capacity Mounted
on
/dev/dsk/c0t0d0s0 96019 80671 5748 94%
/
/dev/dsk/c0t0d0s3 424798 242886 139442 64%
/usr
/proc 0 0 0
0% /proc
fd 0 0
0 0% /dev/fd
/dev/dsk/c0t1d0s1 2466551 9 2219892 1%
/free
/dev/dsk/c0t0d0s5 961257 231471 633666 27%
/opt
/dev/dsk/c0t1d0s0 1475948 210756 1117602 16%
/usr/OV
/dev/dsk/c0t0d0s4 1475948 126654 1201704 10%
/usr/local/Tivoli
swap 1401872 152 1401720
1% /tmp
(See attached file: core)
---------------------- Encaminhado por Fabiana Schurhaus/Telesc em 24/06/99
14:22 ---------------------------
James Shanks <James_Shanks AT TIVOLI DOT COM> em 25/10/99 11:48:15
Favor responder a Discussion of IBM NetView and POLYCENTER Manager on NetView
<NV-L AT UCSBVM.UCSB DOT EDU>
Para: NV-L AT UCSBVM.UCSB DOT EDU
cc: (bcc: Fabiana Schurhaus/Telesc)
Assunto: Re: ovwdb daemon
Well, ovwdb is more fundamental than ovtopmd, so if he's not running, no
important daemon will start.
But trying to start the NetView GUI with the ovw command when you have daemon
problems is doomed to fail. The ovw command does not check for any daemons not
running. You must instead use the netview command.
I have lost the thread of your original append since you are not keeping them
together. But I believe you said that the original problem was that the daemons
would not start at reboot automatically but that you could start them later from
the command line. Correct?
Is there a core file for ovwdb in /usr/OV/PD/cores? If so, I would call
Support. If not, there are a few things I can suggest you might try.
(1) After a reboot failure, just go to the command line as root and type in
/usr/OV/bin/ovwdb and see how long it takes him to start and what error
messages, if any, you get in the command window. That may give you a clue. If
it takes much longer than 2 minutes to start ovwdb, then I would be sure to do
both (2) and (3) below. They may help in any case. Also format the nettl
logs :
netfmt -f nettl.LOG00 > "myfile0"
netfmt -f nettl.LOG01 > "myfile1"
and see if there are clues in there. You can also look in
/usr/OV/tmp/netnmrc.ovstart but I don't think there will be anything in there
you don't already know.
(2) Once you get ovwdb and ovtopmd running, do some database clean up. First
run ovwls and see how many maps you have, and then run (a) ovmapcount -a and
(b) ovtopofix -A (if you have only one map) or ovtopofix -a if you have more
than one.
(3) Run ovobjprint -S and get a count of objects in your database. Then make
sure that the ovwdb cache size is comfortably larger than this by about 20%.
You can see what your current cache is by doing ps -ef | grep ovwdb and seeing
what value is attached to the "-n" parameter. You can raise this value by
using the Tivoli Framework GUI. Click on the policy region to bring up the
NetView server icon, then use the right mouse button to get the context menu for
that. Then choose "Configure --> Set options for daemons --> Set options for
topology, discovery, and database daemons --> set options for ovwdb.
You have said how long this problem has been occurring, but according to my
notes, some folks on Solaris have a problem with the answer book. You didn't
move this from port 8888 to 9999, did you? That what ovwdb expects to use.
>If you run TME10 3.6 and Netview 5.1 on Solaris 2.6 with AnswerBook 2
>installed, port 8888 is occupied by AB2, and Netview would not start. (this
>problem will be logged in /tmp/NVport ---something), stop ab2, and
>reconfigure it to use other ports.
The only other thing I can think of is that there is something different about
your environment as root that is missing at the time netnmrc is run, but at this
point I cannot tell you what.
James Shanks
Tivoli (NetView for UNIX) L3 Support
fabiana AT telesc.com DOT br on 10/25/99 08:59:18 AM
To: james_shanks AT tivoli DOT com
cc: (bcc: James Shanks/Tivoli Systems)
Subject: Re: ovwdb daemon
I think that the daemons are started with /etc/init.d/netnmrc and the ovwdb
daemon
don't start:
ovwdb = state: NOT_RUNNING and exit status exited on signal 9
and
ovtopmd, netmon, ems_sieve_agent, ems_log_agent, nvcold, snmpCollect are
with state: UNSTARTABLE
If I try to start the netview with the command : /usr/OV/bin/ovw the interface
inicialize and after
terminated because shows the message : ovw: cannot to open the database.
(See attached file: netnmrc.ovstart)
Your subject says "ovwdb" but your posting makes no mention of it.
In your posting, all the daemons you mention require ovtopmd to be running as a
pre-cursor to themselves. So what does ovstatus ovtopmd say was the last error
message from ovtopmd? If it is an exit 255, then the problem is most likely
that the Tivoli environment is not being sourced before the daemons are being
started. ovtopmd requires the Tivoli libraries in his path.
How is NetView started at reboot in your environment?
James Shanks
Tivoli (NetView for UNIX) L3 Support
____________________________________________
Fabiana Schurhaus
TELESC - Telecomunica
ções de Santa Catarina S/A
Tele Centro Sul
Florianópolis - SC < http://www.telesc.com.br >
E-mail: fabiana AT telesc.com DOT br
Voice : 55-48-231-2812 FAX : 55-48-231-2611
___________________________________________
netnmrc.ovstart
Description: Binary data
core
Description: Binary data
|