| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 1318.1 | a good problem | ENUF::GASSMAN | | Wed Aug 07 1991 12:22 | 10 |
| This problem will probably be one of the ones displayed at Interop, in
the SNMP 'bad-boy' section. There is a bakeoff of sorts going on, to
see which managers can find which problems. The only tool that I've
seen do this to any degree was the vitalink bridge manager, which could
detect nodes with the same ethernet address on both sides of a vitalink
bridge. This would work with DECnet, cause a DECnet address effects
the ethernet address, but in IP, there is no relationship. I don't
think the problem has been solved - but at interop, we'll see.
bill
|
| 1318.2 | What is the plan?
| HGSWS1::SUSANCHAN | | Wed Aug 21 1991 07:36 | 6 |
|
Is there any plan on detecting duplicate either IP address or DECnet node
address through DECmcc?
Thanks
Susan
|
| 1318.3 | | BSYBEE::EGOLF | John C. Egolf LKG2-2/T02 x226-7874 | Wed Aug 21 1991 08:50 | 8 |
| Susan,
The answer is "yes", but we don't have a timeframe or specific
functionality.
This is one of those things that "we gotta do", it just hasn't
been nailed down as to who and when. We have work with others
within Digital to get the function performed.
|
| 1318.4 | Not solved - yet | ENUF::GASSMAN | | Thu Aug 22 1991 16:49 | 11 |
| Someone also has to figure out HOW!? This is a problem that plagues
both DECnet and IP network managers. Most approaches at solving the
problem resolves around poking at smart lan probes and bridges to see
where the device may be hiding. Someone is going to make a bundle if
they figure out the solution first, and get it to market before telling
everyone how it's done. HP has a new tool out called the "network
advisor" that will find things like IP nodes that have the wrong
broadcast address and such, but I haven't seen them brag about the
duplicate address problem yet. The price is around $20K for the tool,
hinting that the 'symptom finder and problem solving' applications can
command a higher price than 'statistics and threshold' systems.
|
| 1318.5 | Another opinion | YOSMTE::SITZ_GL | | Mon Aug 26 1991 12:51 | 17 |
| re .3 & .4
John, Bill,
I have two types of customers who look at DECmcc: Type one are
"management" people who are not very technical. Type two are "network
management" people who are usually very technical. 100% of the "type
two" people ask about duplicate address detection. Nearly half of the
"type one" people ask the same question.
Opinion: Customers want/need this very badly. We should have a
definate plan and plenty of resources to develop it. Bill is correct,
customers will pay a lot of $$$ for this functionality.
Regards,
Glen R.
|
| 1318.6 | We know how to detect duplicates | CAPN::SYLOR | Architect = Buzzword Generator | Thu Sep 12 1991 22:43 | 11 |
| Re .4:
Yes, we know how :-)
Seriously, we spent a fair amount of time on this in the design of
Phase V. A pair of DECnet nodes with the same address or name can be
detected using SYSID, and/or some of the events a node emits. I won't give
the details here. When someone wants to implement/design code to do
this, contact me.
Mark
|
| 1318.7 | | CADSYS::LEMONS | And we thank you for your support. | Thu Sep 10 1992 15:49 | 6 |
| I really don't want to write the code myself. What is the status of a 'real'
software engineer or programmer doing the job, such that I can be sold to a
customer?
Thanks!
tl
|
| 1318.8 | see 2861.* | TOOK::MCPHERSON | Life is hard. Play short. | Fri Sep 11 1992 16:00 | 2 |
| DCL 'starter' procedure in 2861.5
|