Hernan -
It is highly unlikely that you will be able to resolve cores in ovwdb or ovtopmd
by "fixing" your database, unless the cores started to happen after you made
changes to it. Normally, the only way to resolve a core is to contact
Support, and run the /usr/OV/service/readcore utility provided on AIX (which
requires that you have the AIX debugger dbx installed) so you can give them the
stack trace it provides. A core indicates a problem so fundamental that the
code in the failing daemon did not anticipate it. And ovtopmd should not core
after you run ovtopofix.
The messages you are getting indicate that you have symbols on the map which are
incorrect. When ipmap starts, he synchronizes with ovtopmd, getting all the
changes since the map was last closed. Apparently, ipmap gave him some object
ids which ipmap cannot resolve in ovwdb. This by itself is not serious, but it
does mean that you have database inconsistencies that should be resolved.
This is the sort of thing that ovtopofix was created for (and ovtopofix is what
"Resolve database inconsistencies" runs). To solve the missing interfaces
problem, if that were all you had to do, after ovtopofix failed to get rid of
them, you would try to locate the symbols on the map and then delete them,
letting netmon re-discover the objects if they were really still there.
But the fact that you have cores after running ovtopofix indicates that you have
fundamental problems with your current databases. I'd say you have two options.
The first is to call Support and enlist their aid with your cores and running
some cleanup procedures. This will take time, so keep backups. But ultimately,
it may not work. Some problems are so bad that no utility can resolve them.
Then you must use SMIT "Maintain" and remove your databases and start over. The
other alternative you have at this point is to not attempt a piecemeal solution
but to wipe out your databases now and rediscover the network so that they get
rebuilt correctly. Not a very attractive choice, I know, but it is my best
answer. Perhaps someone else has another.
James Shanks
Tivoli (NetView for UNIX and NT) L3 Support
"Hernan Dario Villegas Londono" <hvillega@bancolombia.com.co> on 04/05/2000
10:12:22 AM
Please respond to IBM NetView Discussion <nv-l@tkg.com>
To: IBM NetView Discussion <nv-l@tkg.com>
cc: (bcc: James Shanks/Tivoli Systems)
Subject: Re: [NV-L] Netview Error
Yes James, I am trying to solve something ; I saw these messages and I wanted
to know if there were problems with them.
My problem is that Netview has been malfunctioning and it has shown cores in
the /usr/OV/PD/cores/ovwdb and in the /usr/OV/PD/cores/ovtopmd.
3 weeks ago I received the ovwdb core and I compressed the database ( Leslie
sent me a procedure to do this) and ran the " Resolve database inconsistences"
and Netview remained stable.
Yesterday I saw the ovtopmd core, and after this, the GUI started with
problems with the collections and the manager submap became "blue" . In
addition, it does not allow me delete or add any map ( normally I used 2 maps).
Probably there are more problems, but I restored a backup and is stable again.
I am worry that this happen again.
Do you know if the problem is because another reason ?
Today, after running the GUI , the nettl.log00, showed a lot of messages like
this :
Initializing formatter....
Initialization done.
************************************ NetView *******************************@#%
Timestamp : Wed Apr 05 2000 07:44:36.998264
Process ID : 17770 Subsystem : OVW
User ID ( UID ) : 0 Log Class : ERROR
Device ID : -1 Path ID : -1
Connection ID : -1 Log Instance : 0
Host Dropped Msgs : 59 Host Dropped Data: 19352
Software : /usr/OV/bin/ipmap
Hostname : antioquia.bancolombia
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
InterfaceSymbol::getNetId(): getNodeByIfID() failed for interface 277678.
************************************ NetView *******************************@#%
Timestamp : Wed Apr 05 2000 07:44:36.998922
Process ID : 17770 Subsystem : OVW
User ID ( UID ) : 0 Log Class : ERROR
Device ID : -1 Path ID : -1
Connection ID : -1 Log Instance : 0
Software : /usr/OV/bin/ipmap
Hostname : antioquia.bancolombia
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
InterfaceSymbol::getNetId(): getNodeByIfID() failed for interface 277679.
************************************ NetView *******************************@#%
Timestamp : Wed Apr 05 2000 07:44:36.999423
Process ID : 17770 Subsystem : OVW
User ID ( UID ) : 0 Log Class : ERROR
Device ID : -1 Path ID : -1
Connection ID : -1 Log Instance : 0
Software : /usr/OV/bin/ipmap
Hostname : antioquia.bancolombia
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
InterfaceSymbol::getNetId(): getNodeByIfID() failed for interface 277680.
Thanks for some help.
Hernan Dario Villegas.
James_Shanks@tivoli.com con fecha 05/04/2000 08:33:34
Por favor, responda a IBM NetView Discussion <nv-l@tkg.com>
Destinatarios: IBM NetView Discussion <nv-l@tkg.com>
CC: (cci: Hernan Dario Villegas Londono/BIC)
Asunto: Re: [NV-L] Netview Error
Yup, I know what it means. And it is not serious. Are you just reading the
nettl looking for problems or are you trying to solve something?
ovtopmd is a trap receiver. He registers for traps from trapd. The external
(non-NetView) ones he just ignores. He has nothing to do for them. For NetVew
ones, he has an internal table which tells him what to do with that trap. Many
internal (NetView) traps are just ignored too. ovtopmd only needs those which
mean he needs to update the topology database. All that this error means is
that he got an internal (NetView) trap that isn't in his table. So apparently
we added a trap some time ago and did not update ovtopmd's internal table. But
it does not affect his processing, since any new trap we added that he doesn't
know about, he doesn't need anyway. So if we eventually update the table, it
will be to tell him to ignore it anyway. Nothing to worry about.
If you do have some problem you are trying to solve, this probably has nothing
to do with it.
James Shanks
Tivoli (NetView for UNIX and NT) L3 Support
"Hernan Dario Villegas Londono" <hvillega@bancolombia.com.co> on 04/04/2000
06:00:14 PM
Please respond to IBM NetView Discussion <nv-l@tkg.com>
To: nv-l@tkg.com
cc: (bcc: James Shanks/Tivoli Systems)
Subject: [NV-L] Netview Error
Hi all , Netview 5.1.2 , AIX 4.3.2 , H50 , 1 GB RAM.
I have seen in the nettl.LOG00 the next messages :
*******************************
Timestamp : Thu Mar 30 2000 17:05:44.110901
Process ID : 32704 Subsystem : OVW
User ID ( UID ) : 0 Log Class : ERROR
Device ID : -1 Path ID : -1
Connection ID : -1 Log Instance : 0
Software : /usr/OV/bin/ipmap
Hostname : antioquia.bancolombia
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
topoEventHandler(): topoEventDispatch() failed. Return code = -2: Internal
error -- could not decode event.
************************************ NetView *******************************
Timestamp : Thu Mar 30 2000 17:07:02.632152
Process ID : 29664 Subsystem : COLLECTION
User ID ( UID ) : 203 Log Class : ERROR
Device ID : -1 Path ID : -1
Connection ID : -1 Log Instance : 0
Software : /usr/OV/bin/collmap
Hostname : antioquia.bancolombia
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
collmap: topoEventHandler(): topoEventDispatch failed. Return code = -2:
Internal error -- could not decode event.
Does anybody know what do these messages mean ?
Thanks,
Hernan D. Villegas
_________________________________________________________________________
NV-L List information (unsubscribing, policies, posting, digest version,
searchable archives): http://www.tkg.com/nv-l
_________________________________________________________________________
NV-L List information (unsubscribing, policies, posting, digest version,
searchable archives): http://www.tkg.com/nv-l
_________________________________________________________________________
NV-L List information (unsubscribing, policies, posting, digest version,
searchable archives): http://www.tkg.com/nv-l
|