🚀 go-pugleaf

Brian Kramer (bjkramer@pluto.njcc.com) wrote:

: Here's the problem:

: I've got 50 lines coming into my site.  I've got multitech Managed Rack 
: 28.8Ks and motorola power modems.

[stuff deleted]

: I talk to the engineer in charge of my central office and he says they
: can't do anything because they are only required to support 9600 baud.
: Which in my mind in a bunch of bull.  This isn't the 80's anymore Bell 
: Atlantic, wake up.

Typical BA...

: They have suggested 
: 1) I purchase $8K channel banks for each 24 lines and pay an extra 
: $300/mo to bring the lines in on a T1. 

Ditto.

: 2) Try "custoflex" lines which they say are their highest quality.  These 
: are only $5 more a month.  I'm trying a few of these to see what happens.

Custoflex is just another name for Centrex service.  As far as I
know, the line quality does not change.

Regards,

Chris

-- 
Christopher Mauritz         | For info on internet access:
ritz@mordor.com             | finger/mail info@ritz.mordor.com OR
Mordor International        | http://www.mordor.com/
201/212/718 internet access | Modem: (201)433-7343,(212)843-3451

Better watch out! Often the degradation of our dial-up lines is the
cause of "not the lines" that worked before, but the trunking network.

RBOC's will install compression hardware with voice compression, fax
bypass, and now... an attempt at dial-up data bypass in the form of
voice/fax/data compression concentrators. Unfortunately, although fax
bypass is now a relatively mature technology when layered with low
rate voice channel encoding (ATC/ADPCM/LD-CELP, etc...) the data
bypass end is NOT stable. If we're having dial-up data problems
en-masse to a POP all of the sudden, the metallic plant hasn't
suddenly becom bad! Something else is wrong.

1. You're beta testing your telco's attempts at compression; without
cost benefit.

2. Once your telco gets compression working properly (at your cost...
se 1) above). you will spend the same for connectivity while their
costs drop dramatically!

See your Public Service Commission... they're usually politicians with
MBA's on the take... but the hardware changes should be accesible for
your local telco should be available through your local PSC... if you
live long enough.

D.R.
Jakarta


barrett@freedomnet.com (Victor Escobar) wrote:

>bjkramer@pluto.njcc.com (Brian Kramer) wrote:

>>1) Dropped connections
>>2) Slow connect speeds, 19.2-21.6K on average for 28.8K modems
>>   4800 and 9600 occasionally with 14.4K modems  The fact that 14.4K 
>>   modems can't connect at 14.4K makes me convinced there is a problem.
>>   I've never seen that before until they made some changes with my wiring.
>>3) Unable to connect at all after retraining
>Uh-huh!  I used to work at an ISP and we had the same problem:
>terribly noisy lines.  It took 5 calls to the State Corporation
>Commission to get them off their arses.  Call NJ's equivalent and keep
>filing complaints till your requests are fulfilled.  The customer is
>always right.  Don't let them tell you any differently.

>>They have suggested 
>>1) I purchase $8K channel banks for each 24 lines and pay an extra 
>>$300/mo to bring the lines in on a T1. 
>>2) Try "custoflex" lines which they say are their highest quality.  These 
>>are only $5 more a month.  I'm trying a few of these to see what happens.
>>3) I am also asking them to check for bridge traps.
>Yup.  They suggested either a T1 or Centrex lines, which IMHO are shit
>and not worth the extra money.  We don't *need* call-forwarding for
>data lines.

>>1) What can I do to get the phone co to do something.
>>2) What can I do to get some hard data to show them the problem.  I'm 
>>willing to purchase some testing equipment.

>#2 isn't necessary.  Like I said, pursue legal channels.  Threaten to
>sue--and mean it!  It's amazing how otherwise lazy BA people suck up
>right to you when the SCC opens up a formal investigation.

>3) I am also asking them to check for bridge traps.

Definitely have them check for this.  We had some bridge tap on a two pair of
wires that was being used by us for a Frame Relay connection.  It was enough to
cause the signal to degrade a couple of times per day and cause a reset of the
connection - took about two minutes to come back up every time.  The telco
(GTE) finally sent someone out at 3 am with a $125,000 UNIX based pattern
testing portable computer to send simulated traffic down the circuit.  Finally
one of the test patterns reproducibly caused the link to die.  The tester was a
stickler who wasn't going to leave until he had tried every possible pattern.
Even though you're not doing the same application, make sure whoever runs tests
on your lines runs every conceivable test...you never know which one will be
significant ahead of time!!

JPF