T.R | Title | User | Personal Name | Date | Lines |
---|
898.1 | we have the same requirements ! | MFRNW1::SCHUSTER | Karl Schuster @UFC OMS/Network Services Munich | Wed Apr 13 1994 03:49 | 8 |
| After my first experience a few weeks ago, and after the first customer
presentations, exactly the same questions/requirements as described in
.0 have been asked.
It is very important to get these features in the future.
Karl Schuster
OMS Munich
|
898.2 | What about Domains | WELSWS::GREENR | Curiosity Killed The Cat | Thu Apr 14 1994 06:59 | 9 |
| My understanding was that by using the "Domain" feature within
Probewatch you could break out LAN utilisation by protocol type. ie you
could set up seperate domains for each protocol type, this is one of
the unique RMON+ capabilities of the Frontier software because rmon
itself does not provide a breakout by protocol type definition.
regs
ross
|
898.3 | But no distribution chart... | WARNUT::BANKST | Network Mercenary | Thu Apr 14 1994 13:09 | 21 |
| Hi Ross
Yes, this is true, but what I (and others) would love to see is a chart
like the one for packet distribution, but for all protocols or selected
protocols. Domains let you select a variety of up to 16 (?) protocols
and see the overall traffic that they generate, but NOT how much each
one generates.
To do this you can run several domains concurrently, and see agraph for
EACH protocol in the seperate domains, but you can't as far as I can
see generate a graph as described above. Lets put it on the wish list!
I used the protocol decode today for the first time, very easy and
quite usefull!
Now how do I get the reports I mentioned in .0 ?? Surely not time to
RTFM? :-)
Cheers
Tim
|
898.4 | PROBEwatch answers (some anyway)... | NAC::FORREST | | Thu Apr 14 1994 18:46 | 47 |
| >One thing though that a number of people who also looked at it commented on
>was there appears to be no way to monitor the top ten protocols on the LAN.
>You can get top ten talkers and listeners but not the noisiest protocols. Is
>that correct?
Ross is correct, you can guess at which protocols are in use and
set up domains for each. You can then see the full RMON statistics,
Host Table, Traffic Matrix etc for each protocol. Only the Frontier
software does this. They also now have working the automatically
generated chart of the top n protocols. We should get this in
PROBEwatch for OSF, perhaps earlier in PROBEwatch for Windows.
>Also, the reporting, is there a way of leaving the reporting running as a
>backgroung process so you can log out the workstation, and have it capture
>data for you on a long term basis?
Yes, the probe will keep long term history - you can see this on
the Segment Zoom screen. Check out the new reporting in the T1.1.3
ULTRIX kit (see NAC::MANAGEMENT:RMON_PROBEWATCH_POINTER.TXT).
>If there is a way to do the above, can it provide you with data in a form
>that you can extract and use in a spreadsheet for example?
I think the T1.1.3 version may let you export in delimited ascii.
If not, be patient, it is coming.
>Is there a way to have the displays show you node names - DECnet, LAT server,
>IP names etc like you could in LTM with a table of names/addresses?
Yes, you can edit in names today. There should be a Names Editor
under the Tools menu on ULTRIX (not on PROBEwatch for Windows yet).
I'll try and post a how-to here later. I think that discovery will
have an agent decode packets to find IP address to MAC address
matches.
>Will it autodiscover just IP names or DECnet as well? And with IP, is there
>a way to hook PROBEwatch into BIND to allow dynamic node name lookup?
It will discover IP names in the /etc/hosts file. It does an
automatic conversion of MAC address to DECnet Phase IV address, but
it does not do DECnet names yet. I don't believe that it hooks into
BIND, but I've seen someone perform a 2 minute operation to take
a name server database and put it on the local host , where
PROBEwatch could find it. I'm not literate enough to know how to
do this.
|
898.5 | Availability on AOSF? | ARRODS::NADER | Reorganising the reorganisation again? | Thu Apr 21 1994 08:42 | 14 |
|
This may not be the right forum for this question but...
In view that we are now selling Alpha/OSF, what are the plans for the
port of PROBEwatch onto this platform. The questions that I need
critically answered are:
o When is it going to be available on Alpha/OSF?
o How much?
o As an interim measure, will the software run on DECmigrate?
Appreciate any info.
Nader.
|
898.6 | PROBEwatch for OSF/1 AXP - some answers.. | NAC::FORREST | | Fri Apr 29 1994 21:53 | 12 |
|
Nader, this is the right forum.
o When is it going to be available on Alpha/OSF?
> check NAC::MANAGEMENT:STATUS.TXT
o How much?
> $3,500 most likely
o As an interim measure, will the software run on DECmigrate?
> don't know - anyone out there with an OSF/1 AXP machine want to try it?
|
898.7 | Seems to work OK | KAOFS::BOIVIN | Moi, j'viens du nord! | Thu May 05 1994 14:19 | 10 |
| Good day,
I've mx'ed (using DECmigrate V1.2) the PROBEwatch V2.5.3 (PWULT110-T3.tar) kit
on my DEC 4000-610, OSF/1 V2.0 (rev 240). My display is to a color VXT 2000.
It seems to work OK with an occasional X Toolkit Warning error.
Hope this helps,
Ed
|
898.8 | So why can't we test it a bit & sell it? | CUJO::HILL | Dan Hill-NetMgtConsult-Denver-553-3624 | Thu May 26 1994 02:52 | 6 |
| This is quite good news.
Why can't some engineering resources be used to test this port and
make it available???
-Dan
|
898.9 | | QUIVER::SLAWRENCE | | Thu Jun 02 1994 18:53 | 11 |
| Actually, one of the most difficult resource constraints we have in the
DEChub program is testing - if you look at the (increasing) number of
products and features we have and then try to imagine all the possible
combinations you'll realize that testing even a fraction of them is a
huge job. Each new product or feature _multiplies_ that number by
something (not adds).
Obviously, we can't stop building new stuff because of that, but it is
certainly true that we've decided not to support some things at least
in part because we could not test them well enough (even though they
were basically done).
|