Conference netcad::hub_mgnt

    
    
    Well don't blame Hubwatch for everything ;^).  Number 1 belongs to me.
    Number 2 and 3 I suspect also belongs to me and possibly 4. Please send 
    me the snmp commands that you are using for Number 2,3, 4 and I
    will have a look see.
    
    							Rene'
    
    							GIGAswitch/FDDI
    							firmware team
    

    
    Otto,
    
       As item 1 states HUBwatch does not show the proper PMD on the
       Front Panel.
    
       For item 2, as pointed out in the previous reply, an SNMP trace will 
       help track this down.
    
       For item 3, is this the bad FGL-4 being refered to or any FGL-4? 
    
       Could you explain item 4 further? Does this refer to FGL-4 in
       general or the defective FGL-4? Also what particular
       views/fields/LEDs were incorrect?
    
       Since HUBwatch uses FDDI related MIBs to determine the LED states on
       an FGL-4, it is not possible to obtain LED states given by
       diagnostic FW. I guess the answer to item 5 depends on the extent of
       the linecards problems and whether FDDI is running or if the
       linecard is halted in diagnostics.
    
       I believe that HUBwatch does show the Port Type correctly (item 6), 
       however a situation may exist where this is not true (ie a bug). An SNMP
       trace will help narrow this down.
    
       HUBwatch does not show DNA, but does show UNA on the Linecard
       Summary View (item 7). We could add DNA to the wish list.
    
    
    Regards,
    Bob

    P.S. There is an entire family of bugs in the FGL-4 firmware which are
    caused by confusion of port numbers in the FGL-4 SNMP firmware.
    
    Re: .2 - please indicate exactly which steps you are taking in Hubwatch
    
    Re: .4 & .5 - which displays in Hubwatch are you using?  If the module
    fails diagnostics, it should indicate the problem in the Operatinal
    Status field of the System Summary display.
    
    MDL

Geez, Bob!

You must live right.  Someone reports a "HUBwatch" problem and two
GIGAswitch people step in to claim it as their problem???

What's your secret?   8-{)}

    
    Just lucky I guess:-)
    
    bd

We used HUBwatch for Windows 4.1 and the IRIS Analyzer to obtain this
traces. All traces are available in File MUNICH::EC_TRACE.BCK. You will
also find a more complete SNMP database for GIGAswitch variables. I
compiled this with all MIBs used by the GIGAswitch/FDDI V1.2. This is
for GIGAswitch Firmware 2.2. Copy PD_GIGA.DAT to your Iris directory as
PD_SNMP.DAT.

Number 1. Trace PORT_PMD.DWC/TXT

As mentioned in previous reply SNMP variable ePORTPMDType.n.1 returns
an integer value of pmdMM(2) instead of pmdUTP(7) when a TP-PMD is installed.

This is probably a FGL-4 Firmware bug.

Number 2. Trace PORT_OFF.DWC/TXT PORT_ON.DWC/TXT

From the FDDI/GIGAswitch Line Card Summary window the first port was
disabled. An SNMP set for variable {snmpFddiPORTAction.29.1} value
disable(4) is send to the GIGAswitch. After this, Hubwatch reads back all
Line card parameters with GetNext command. The returned variable
{snmpFddiPORTConnectState.n.1} for ports 29,30,31 and 32 shows a value of
disabled(1). Only port 29 should show this status. Ports 30,31 and 32
should still be (e.g.) connecting(2) (no FDDI station connected there).
After port 29 is enabled again all other ports on this card show
{snmpFddiPORTConnectState.n.1} enabled(2).

This is probably a FGL-4 Firmware bug.

Number 3.  Also see Number 6

This one is ment to be the LED indication on the HUBwatch Main Display.
The right and left PMD LEDs are on the GIGAswitch board are as described
in V2.2 Release notes. But they don't light the same way on the 
HUBwatch display as on the physical Box

This is probably a HUBwatch bug because the SNMP traces show correct A/B and
M/S port snmpFddiPORTPCType.n.0's.


Number 4. and 5.

Re: .3 Yes the status is displayed in the system summary display. What the
customer wanted is a trap on Linecard status changes. He had a problem that
the card was continuously running selftest. He wanted a trap whenever a selftest
is initiated. This may be good for the wishlist.

Number 6. Trace PORT_M.DWC/TXT 

If Port 30 (second port on FGL-4) is managed to M/S the snmpFddiPORTPCType.n.0
on all Ports are reported with a value=s(3) or m(4).

This is probably a FGL-4 Firmware bug.

Hubwatch does not indicate the correct LED Status for PMDs on the Main display.

Number 7.

DNA is also mentioned for the wishlist

One more. HUBwatch for Windows 4.1 has a problem displaying the number of
ports in a slot as well as for the modules. It also does not display the
seven segment indicator.

Regards		Peter
    

    
    Peter,
    
    Thanks very much for the feedback.
    
    Number 3:
    
    There is a problem with the right PMD LED in HUBwatch. It is never lit.
    
    Number 7:
    
    Could you please explain this problem ("problem displaying the number
    of ports...") a bit further. Which HUBwatch view is this on?
    
    The seven segment indicator is not currently supported under Windows.
    We will add that to the the wish list items.
    
    Thanks again,
    Bob
    

    
    >>>1. Hubwatch does not correctly show the type of PMD ( TP is seen as Fiber)
    
    >>>2. Switching off one Port on a FGL-4 Card switches off all Ports on that card
    
    >>>3. Neither the Port State  nor the PMD State is shown correctly
    
    Okay these things are now fixed and will be placed in the V3.0 path of 
    GIGAswitch/FDDI firmware. Two bugs were found, one in FGL4 code and
    one in SCP code, so obviously you will need both. As it stands now
    you will only see the status/PMD type/switching off port for Port 2 
    of the FGL4 no matter which port of that card you select.
    
    								Cheers,
    
    								 Rene'