Stigers, Greg ~ AND wrote:
>
> If someone with some experience DTC TelnetAccess could reply to me via
> email, I am trying to make some sense of this from just the LaserROM.
> I gather that one can have one (and only one?) Telnet Access Card
> (TAC) in a DTC48.

Unless recently changed, correct as I recall.  And on a TAC on a DTC48
you are limited to 40 concurrent telnet sessions; above that and you
will get 'connection refused'.

> Assuming the desire to telnet to the 3K across a WAN, could the DTC
> be local to the users on the other side of the WAN, and how well
> does that work? Or, if the DTC should be more local to the host, and
> the TCP/IP traffic crosses the WAN (which is our area of difficulty,
> poor performance across a congested and perhaps problematic WAN),
> would there be much advantage to using the DTC48?

If the non-3000 side of the WAN is a different subnet, you'll need to
have the DTC on the 3000-side or else bridge over the WAN (even more
traffic, depending on what you have on your LAN).

> Assuming dedicated users, how many user can one DTC48 reasonably
> support? Does one telnet to the IP address of the card or the 3K?

If you telnet to the IP address of the card, you get the DTC user
interface.  You can associate other IP addresses with HP Probe names of
3000 hosts such that telnetting to that IP address does an automatic
connection to the associated 3000.  You never telnet to the IP address
of the 3000 in the case of the DTC.

> How might this compare using with inetd on a 996/200?

If the 40 users on the DTC48 represents your expected telnet load on a
996, I'd say just enable host telnet via inetd.  On a 950/960 that
number is noticeable, but a 996/200 probably won't flinch.  If you do
not have a TAC in your hands, I certainly wouldn't spend the money on
it given your scenario.

You could get routable DTC too, but it requires Openview DTC Manager
stations on each end and another suitable DTC (still more expense).

Jeff Kell <[log in to unmask]>