Ascend Archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: (ASCEND) SNMP on Channelized E1



David Ponzone wrote:
> 
> Here is my problem.
> 
> I have a Chanelized E1 coming into a Max 4000.
> 
> On this E1, I defined 2 groups, group 1 and group 2.
> I have a connection profile (nailed) for each group.

Interesting question for me: Is there any documentation available on
what really happens down in the E1 time slots (B-Channels) when doing
souch grouping ? I've tested a P400E1 against a Max2000 with a cross-
over E1 cable, and when placing all available B-Channels (1..31) into
Nailed mode, I can establish a connection nicely. The resulting band-
with is displayed differently here and there, but that's not the
matter if I know what really happens. When using a PPP profile on it,
is the result PPP framing on the thick chunk, or PPP framing on single
B-Channels ? Or is indeed MP(P) used ? What if I really say MPP and
Nailed/Mpp - do I get inverse multiplexing on 31 Slots or one MPP
channel on a 1984 kbps line ? Where is this documented, if at all ?

(Intention is to drive a E1 leased line [German Telekom D2MS] with
Ascend Equipment - and no, I don't know if there ever will be a
P130 E1, but it would be great...)

BTW, in the above case the configuration of Net/E1 is:

Sig Mode=NONE
Switch Type=DASS 2	(but that won't matter anyway...)
Framing Mode=G.703

and all channels are Nailed and Ch n Prt/Grp=1.

> As the wan interface assigned to a connection profile is not
> fixed, and will change each time the connection goes down and up,
> I can not monitor the ifInOctets and ifOutOctets with SNMP.
> (BTW, for a 256Kbps line, so a group of 4 64Kbps-channels, here
> is the output of 'sh if st' on the Max:
> wan58            ppp     Up     23     64000      1500     125647     104412
> Well, apparently, some problems to get the real bandwidth.
> 
> Here is the output of 'sh users':
> O 235943603 3:0     9:2   64K   PPP         x.x.x.x       customer_profile
> Still some problems to show the bandwidth.
> 
> But if I request the variable callStatusDataRate, I get 256000.
> Well...
> )
> 
> I don't really care about the inconsistencies reported by the
> terminal server, BUT:
> 
> does anyone know a way to get the SNMP variables on a such
> interface, cleanly and definitely (I mean, without reconfiguring
> my SNMP client each time the interface goes down/up :)) ?
> A clean way would be to be able to access groups as if they were interfaces,
> as the group number is the only fixed thing.
> Or the profile name.

That's the normal problem with all network devices that use a dynamic
interface assignment approach, the problem usually is insufficient
intelligence in the NMS which assume interfaces are very static -
which may have been true in the fields of leased lines with transfer
nets in the good old backbone days, but isn't a feasible concept with
dialin/dialout NAS style devices like Ascends.

What you need is a bit logic to go from a profile name to one or
a list of (think of MP(P)) ifIndexes. This is possible by heavily
digging in the vendor MIB, but it's not really simple. IMHO Ascend
should give us a simple and direct mapping that can be traversed
with at most three SNMP GETs.

The last time I checked this out, I had to:

Go into ascend.sessionStatusGroup.sessionStatusTable.sessionStatusEntry
Retrieve all ssnStatusUserName from there

Go into ascend.callStatusGroup.callStatusTable.callStatusEntry
Retrieve all callSessionIndex and callStatusIfIndex from there

Make sense of all the above tables to finnaly get the correct ifIndex
or list of them.

There was even a bug in the vendor MIB that all callStatusIfIndex
values were indeed off by one. I've mailed this, but didn't get
an answer (however this seems to be fixed since late 4.6C, when
the major reorganization of the MIB started).

Probably NetClarity will do this very nicely ;)

Andre.
-- 

Kanther-Line: PGP SSH IDEA MD5 GOST RIPE-MD160 3DES RSA FEAL32 RC4

+-o-+--------------------------------------------------------+-o-+
| o |               \\\- Brain Inside -///                   | o |
| o |                   ^^^^^^^^^^^^^^                       | o |
| o | Andre' Beck (ABPSoft) beck@ibh-dd.de XLink PoP Dresden | o |
+-o-+--------------------------------------------------------+-o-+
++ Ascend Users Mailing List ++
To unsubscribe:	send unsubscribe to ascend-users-request@bungi.com
To get FAQ'd:	<http://www.shore.net/~dreaming/ascend-faq>
or		<ftp://ftp.shore.net/members/dreaming/ascend-faq.txt>


References: