[135] in Enterprise Print Delivery Team
Re: Testing IPM
daemon@ATHENA.MIT.EDU (David F. Lambert)
Mon Mar 13 08:24:41 2000
Message-Id: <10003131326.AA20622@MIT.EDU>
Date: Mon, 13 Mar 00 07:58:24 EST
From: "David F. Lambert" <LAMBERT@mitvma.mit.edu>
To: Enterprise Printing Delivery Project Team <printdel@MIT.EDU>
In-Reply-To: Your message of 09 Mar 2000 14:43:40 -0500
Mike,
I can answer some of your questions...
- Yes, the SAP application does run on Compaq servers. The only connection
remaining between SAP and the mainframe is printing. Monthly GL
print files are directed from SAP to the mainframe for printing.
No functional appl code is executed on the mainframe.
- We do NOT have an InfoPrint 60 (IP60) printer. Neither the 3160 Model 2
nor the 3827 are IPx model printers. We would need a loaner to test
the IP60.
- I think you're right about what they consider "integrated sith SAP"
but it would be nice to confirm this. Other vendors make the same
"lame" statement.
- It would be good if you would confirm the use of the old MCA bus
and old model RS6000s for use with IPM - seems like a step backward.
- We don't want to expand our use of SNA as part of this effort. So,
moving to a model which requires SNA under IP is NOT a direction
we want to move in. Agreed!
- It's unclear as to whether or not we want to continue to use AFP
Formdefs & Pagedefs. CAO has expressed a greater interest in
dumping AFP completely and moving to Postscript. They (customers)
may not fully understand the value and usefulness of electronic
form overlays, etc.
- eNetwork Communications Server is new to me. I assume you're right
about its function, Mike.
- DCE vs non-DCE: I'd only vote for the DCE version if it's easier to
port that version to our Kerberos environment. I would guess the
non-DCE version would be an easier port but what do I know about such
things...
Dave
On 09 Mar 2000 14:43:40 -0500 Mike said:
>Here's my thoughts on this. If nobody objects, I'll turn this into a
>less flamacious version and send it off.
>
>"David F. Lambert" <LAMBERT@mitvma.mit.edu> writes:
>
>> > #1 Accept PS data from SAP running on VM (?) and print on a IP60
>> >TCP/IP attached printer (600 DPI).
>
>The SAP application servers are Digital Unix boxes, right? Does any
>part of SAP actually run on the mainframe?
>
>Is "IP60" another name for one of the IBM printers we already have, or
>a new one we would need to demo?
>
>> > - Question how does the SAP PS data get to IPM? (LPR?) You
>> >would probably feed IPM like you feed the HP's today.
>
>Ugh. So their idea of "SAP integration" is that IPM runs an lpd
>server, and SAP prints to it? That's what the Athena print system
>does right now. If that's all the integration there is, give me some
>protocol specs and a PS->IPDS filter and a couple weeks, and I'll have
>an LPRng filter that does the same thing.
>
>And given that their hype of driving channel-attached printers turns
>out to be less feasible than their advertising likes to point out
>(requires a MCA-bus RS6k, which are not only AIX boxes, they're *old*
>AIX boxes), that's a benefit that's rapidly going out the window.
>
>> > - Installed IP60 requires 10/100BaseTX Ethernet attachment
>> >feature #4165 ($750) field installed.
>
>Do we have this?
>
>> > #2 Continue to print legacy systems output (IPDS) on the IP60 from
>> >VM. (Today this prints via 370 channel connection to the
>> >3827, 3160 and 4245's .)
>> > - This path would be replaced by connecting IPM to VM via the
>> >network and send IPDS data via SNA to IPM's PSF Direct feature.
>
>Waitasec, won't this require SNA support on our routers? I was under
>the impression that Netops really didn't like supporting anything
>other than IP.
>
>> >The PSF Direct feature would convert the datastream from SNA to TCP/IP (low
>> > overhead) and forward to the TCP/IP attached IP60. The 3827
>> >won't play unless a Micro Channel RS/6000 is used and a 370
>> >channel card is purchased for EACH printer requiring a Bus &Tag interface.
>> > The 4245's won't play at all as Line Printers aren't supported
>> >via IPM. 6400's would be an alternative as they are TCP/IP
>> >addressable and IPM capable. (I recommend TCP/IP printers as 370 Channel
>> >cards are $$ and limit the purchase of RS/6000s to less economic
>> >models.)
>
>So. Channel-only printers are useless with IPM unless we get an old,
>used RS6k with MCA channel cards. (Sigh. And I was hoping they would
>just use the same PCI channel cards that they use in the NT version of
>IPM.)
>
>> > - IPM server requires eNetwork Communications Server V5 to
>> >support SNA
>
>"eNetwork Communications Server"? I presume this is a
>buzzword-compliant name for an SNA stack/module for AIX.
>
>> > #3 Accept PS/PCL data targeted to a subset of HP's to be defined on
>> >IPM and used for testing.
>
>Fair enough. All kinds of nasty little screw cases to test here.
>
>> > - This test would be used to demonstrate accounting features of
>> >IPM and use of the to GUI's to define and control the network type
>> >printers and show security features. The datastream would be "passthrough"
>> >and not involve a transform on IPM (low resource consumption) Could
>> >also handle any datastream type but others may involve a
>> >transform. (Higher resource consumption)
>
>Given how little SAP integration they imply, the selling strength of
>IPM will IMHO be its robustness and admin/operator interface.
>
>> > The first Low speed printer feature includes license to drive 20
>> >printers. Additional blocks of low speed printers are added in
>> >groups of 40 up to a max charge of 220 printers. I would recommend the 1st
>> >group (20 HP printers) for our testing based on what I understood
>> >yesterday. Realizing this number could grow to 500+ in
>> >production.
>
>So we pay by the printer? Joy, joy.
>
>> > #4 Would you want to generate AFP resources (Pagedefs and Formdefs)
>> >on the AIX platform?
>> >
>> > Page Printing Formatting Aid (PPFA) is available as an option for
>> >IPM.
>
>I'd guess that depends on what we're doing. Presumably we would, and
>it sounds like a decent resource to have.
>
>> > #5 Do we want to look at DCE during this test period?
>
>No. DCE is not useful to us.
>
>> >Have I missed anything? Please send me a note with additions and/or
>> >subtractions and I will modify!
>> >
>> >
>> >Current thoughts on IPM server configuration:
>> >
>> > 5745-E42 InfoPrint Manager V3.2 (beta not a guarantee)
>> > Base Code
>> > Client Code
>> > 1st Low Speed attachment feature (20 low speed printers
>> ><100ppm)
>> > PPFA
>> > 5765-D20 eNetwork Communications Server for AIX V5.0 (SNA for AIX)
>> > 5639-I35 Distributed Communications Environment
>
>No DCE.
>
>> >Note: MIT to locate an RS/6000 capable of driving test configuration. (May
>> >borrow my 43P but it offers limited perf.)
>
>I guess we're working on this. E40 doesn't have one; I have a very
>old one sitting at home (the kind that was in Athena clusters around
>1993-1994).
>
>For the record, I'm not even willing to consider IPM as a replacement
>for the Athena printing system until it speaks Kerberos.
>
>-mike
