| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 3389.1 | reset before restore | SLINK::HOOD | Your bad news bear | Fri Mar 22 1996 10:40 | 16 |
| One possibility... Version 1 of Resilience does not do resets to modules
and hubs before doing a restore. The documentation recommends doing a
factory reset first.
The reason Resilience (soon to be called "clearVISN Recovery Manager) does
not do a reset first is so that you can restore just specific things (for
example, just restore address filters).
Version 1.1 is just about to ship, but I'll mention it to the folks doing
2.0 to maybe add a factory-reset button for restore operations.
Anyhow, if this isn't your problem, let me know.
Tom Hood
formerly Resilience designer
|
| 3389.2 | Will ip/trap address be restored? | SNOFS1::63496::CHIUANDREW | | Sun Mar 24 1996 17:57 | 11 |
| re .1
Thanks for your quick response!
Do you mean we have to do the factory reset before the restoral, if I
perform the factory reset on a module, will ClearViSN still able to
restore the IP address/trap/community back to the module? Same question
for DEChub900 MAM?
thanks in advance for help!
Andrew Chiu
Network services Sydney
|
| 3389.3 | Manually. | SLINK::HOOD | Your bad news bear | Mon Mar 25 1996 09:32 | 4 |
| No. Resilience only talks SNMP. So, If you do a factory reset, you
will need to manually re-enter the IP info.
Tom
|
| 3389.4 | Thanks! | SLINK::HOOD | Your bad news bear | Mon Mar 25 1996 09:34 | 4 |
| And *thank you* for raising the question in .2 !!! Adding a reset to
Recovery Manager would not work at all!
Tom
|
| 3389.5 | Is it right to say it will take a while? | SNOFS1::63496::CHIUANDREW | | Mon Mar 25 1996 18:05 | 24 |
| re .3/.4
Thanks again for help!
Just want to clarify the procedure to restore the configuration of a
DEchub 900 and its modules (just in case the HUB and modules died).
1) First we have to backup the HUB configuration (MAM and modules) to
ClearVISN.
2) When the HUB and modules ALL DIED, then we use ClearVISN restore
manager to restore the config of that HUB/modules, BUT before we can
restore it, we have to do FACTORY RESET on HUB and all modules, then
re-enter the IP address/SNMP community/trap address for the
HUB/modules.
3) Start to restore the HUB/modules, via ClearVISN Restore Manager.
It seems to take a while to restore the HUB back to operation.
Any comment?
Andrew Chiu
Network Services Sydney
|
| 3389.6 | No!!! Factory reset *should* not be necessary! | SLINK::HOOD | Your bad news bear | Tue Mar 26 1996 02:11 | 9 |
| No, you do not need to factory reset!!! Sorry if I left that
impression. What I meant was that it can help sometimes if
your equipment is in some kind of indeterminate state.
Andrew, can you please mail me as good a problem description of
the error and the hub configuration as you can. You should not
get restore errors.
Tom
|
| 3389.7 | try setting RETRIES to 30 or more | NETCAD::MOWER | | Thu Apr 11 1996 13:12 | 10 |
|
I suggest you increase your RETRIES count, (in your agents file), to 30.
If you are still in a position to test this, please let me know if this
solves your problem.
Carl Mower.
(I wrote the hub900 Recovery Mgr script, and with others, have been
looking into a very similar behavior - this is the best work-around
available at this time).
|
| 3389.8 | still need more information | NETCAD::MOWER | | Wed Apr 17 1996 19:06 | 19 |
|
I got your mail from Tom. While my suggestion in -.1 may help your
situation, I need some more information to verify if it will in fact solve
your problem.
In .0 you listed your configuration as:
1 2 3 4 5 6 7 8
DS900EF empty empty PE900TX PE900TX PE900TX PE900TX DS900EF
| || || || || |
| || || || || |
`------------------------'`--------'`--------'`--------'`--------'
What is the LIGO setting of the modules in slots 1 & 8?
Which slot is your PC (running ClearVISN) connected to?
Which slot is your IP services module for the hub?
What are 5 or so Recovery Mgr status lines just before the error?
Carl.
|
| 3389.9 | More info, hope it helps! | SNOFS1::63496::CHIUANDREW | | Wed Apr 17 1996 20:46 | 43 |
| re .8,
Carl,
DS900EF in slot 8 (let called it EF8), A-port is external FDDI and
B-port connected to BP's FDDI internal ring. DS900EF in slot 1 (let me
called it EF1), A-port connected to BP's internal FDDI ring ad port-B
is connected to external FDDI ring. In fact, EF8's A-port connected to
another DEchub900 with DS900EF (SBEF) as well and EF1's B-port connected to
another DEchub900 via DS900EF (AWEF).
The IBM for our HUB (with 4 PE900TXs, 2 DS900EFs) is EF8 (DS900EF in
slot 8), where EF8 and EF1 has their own IP address as well (i.e.
MAM is 203.5.81.41, EF8 is 203.5.81.42 and EF1 is 203.5.81.43). My
notebook (Hinote running WIN95 with ClearVISN resilience) connected to
EF8's port-6 (4th UTP port) directly.
Before the 'restore', ALL PE900TXs are connected to DS900EFS via
Internal FDDI ring. They are working withiut any problem (I can ping to
their network via any port on anyone PE900TX). Once, I started the
restore (after I removed ALL modules from the HUB, then put them into
the same slot as before, then I started ClearVISN resilience to restore
the config that I backup before). Then PE900TX in slot 5, LED of A-port
is BLACK out and LED of B-port is blinking', when I looked at it via
HUBwatch V4.1.1 (on my PC), there is NOT connection of this PE900TX to
the BP FDDI internal ring, but other PE900TXs (in slot 4, 6 & 7) got
correct BP connections (i.e. they connected to internal FDDI to
DS900EFs), I can ping from those PE900TXs (in slot 4, 6 & 7) when I
connected my PC to UTP ports on them, but cannot ping on PE900TX on
slot 5. I have to manually, disconnected EF1 from the BP, then
reconnect PE5 (PE900TX in slot 5) to BP FDDI, then EF1. After these,
it backed to normal (I meant before we removed the modules from the
HUB). I did not reset any modules during our tests, it seemed to me
that moudles should retain ALL config parameters even you removed from
the HUB. Also I did not reset the MAM, again, it should retain ALL
config parameters (except BP, is it right).
Hope this can give a bit clear about our test. BUt customer already put
this HUB to their network (they have 20+ servers, VAX/AXP/NT..), so we
cannot test it again. But we have to give some sort of 'proof' that
ClearVISN can do backup/restore.
thanks again for looking into the subject matter.
Andrew Chiu - Network services Sydney
|
| 3389.10 | I'd still suggest up RETRIES=30 | NETCAD::MOWER | | Thu Apr 18 1996 15:17 | 38 |
|
Based on the additional information you provided in -.1, I have modified my
picture of your setup. The FDDI ring on the hub backplane is shown below
the 8 modules.
The FDDI ring is drawn the way it is to represent the way the ring actually
connects modules in the hub. The rule is: in the IN, (slot 8 in your case),
then hit all other modules from 1-->8, then out the OUT, (slot 1 in your
case).
I have marked slot 8 as your IPS, (IP services module; you called it IBM).
I have also indicated that your station running Recovery Manager, (the
"nms"), is connected directly to the front of slot 8.
| | nms
| | |
| | V
B A ips
1:900EF 2:empty 3:empty 4:900TX 5:900TX 6:900TX 7:900TX 8:900EF
A A B A B A B A B B
| | | | | | | | | |
| | | | | | | | | |
| | `------' `------' `------' | |
| | | |
`------------------------|----------------------------' |
| |
`-------------------------------------'
I still could use the 5 or so Recovery Mgr status lines just before the
error. Without seeing them, I cannot be sure what the error is, but I
believe my response in .7 is still correct, (up retries to 30).
As to why this [may] solve your problem, we are still working on that.
Stay tuned to this notesfile where I hope to soon enter a new topic
discussing this very problem, (along with other related issues).
Carl.
|