[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference netcad::hub_mgnt

Title:DEChub/HUBwatch/PROBEwatch CONFERENCE
Notice:Firmware -2, Doc -3, Power -4, HW kits -5, firm load -6&7
Moderator:NETCAD::COLELLADT
Created:Wed Nov 13 1991
Last Modified:Fri Jun 06 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:4455
Total number of notes:16761

4286.0. "Recovery Manager Save gets error 89" by CSC32::R_BUCK (Authenticated and assimilated) Mon Mar 17 1997 14:31

    Have a report from a customer concerning clearVISN Recovery Manager. 
    This customer has six differnt DEChubs with a variety of modules.  Two
    of the configurations get an error message when Recovery Manager is run
    to save the current configuration.
    
    Error message contains:
      BRMI GET CALLBACK ... Error 89
    
    Log file shows error with operation:
      Getting Connection table
    
    Details for one of the Hubs that reports a failure message:
    
    MAM Code V4.1.0
    Port 1: Decrepeater 900TM
    Port 2: Decrepeater 900TM
    Port 3: Portswitch 900TP
    Port 4: Decrepeater 900TM
    Port 5: Decrepeater 900TM
    Port 6: Portswitch 900TP
    Port 7: Decrepeater 900TM
    Port 8: Decswitch 900EF
    
    ------------------------------------------------------------
    Any ideas?  More details needed??  So far I have not seen a similar
    error with any of the Hubs onsite here.  Do not have enough modules to
    re-create the exact customer configuration.
    
    Randall Buck
    MCS - Network Support
T.RTitleUserPersonal
Name
DateLines
4286.1Hub900 is in an incoherent stateNETCAD::GILLISTue Mar 18 1997 10:4361
Randall,

WHAT YOUR ERROR MESSAGE TELLS ME:
================================
"brMiGetCallBack(): failed on error 89"
indicates to us as developers that Recovery Manager
failed in function brMiGetCallBack(), and received
a HUB_ERROR message at the time when your Hub900 script tried to
backup the Hub900 connection table.

TRANSLATION:
===========
It sounds like your hub(s) got into an incoherent state, where the
Hub900 backplane connection tables are hosed somehow. The error
you received occurs somewhere in the code that builds the SNMP 
request/response packet, and is not one of the usual SNMP
errors seen by a client (bad value, no response, generrs, etc). 
Here's a few thoughts on how the Hub900 got into an incoherent
state:

1) Did you pop a module in and out quickly, and not wait 15 or so
seconds before kicking off a backup operation? This can cause the 
backplane connections to be in a state of flux (the connection
tables are rebuilt each time a device is removed or inserted), 
and Recovery Manager was reading the Hub900 "Connection table"
when the connection tables were not in steady-state.

2) Was a module removed while in the middle of the backup
operation? Each module is backed up, and then the Hub900
last ... if a module was removed just prior to the backup
of the Hub900, the Hub900 could get into an incoherent state.

3) Have you completed a successful restore of a previously
saved configuration? This is important, because it would mean
that things were in a defined state and were working at one time,
and now they are not (versus "things never worked").

I would try the following to fix your problem:

- try to "reset to current settings" for the Hub900 backplane,
and once the hub comes up, try the backup operation again.

- if the backup still fails, and you have a backup file for the hub that
you were able to restore sucessfully, reset the Hub900 to
Factory Defaults, and try again.
============================================================

In the upcoming clearVISN Recovery Manager V2.0 release,
we've added more user-friendly text to our error messages,
and we are going to run the Hub900 consistency check
at restore time prior to restoring any devices. If we possibly
can, we'll try to run a similar consistency check at backup
time before any modules are backed up ... this would catch
your problem before any devices are backed up, but it's a case
of "sooner than later", and the Hub900 connections are still
hosed regardless.

Keep me posted on your progress.

John Gillis
clearVISN
4286.2See note 4233 for similar errorNETCAD::GILLISTue Mar 18 1997 11:195
See note 4233 for similar error, only it occurred
at Restore-time. 

John Gillis
clearVISN Recovery Manager