[ih] inter IMP hackery [was Recently restored and a small ARPANET was run using simulated IMP hardware, ]

[Alex McKenzie]

 Geoff,
The NOC in the early 1970s was interested in the up/down status of IMPs, lines, and eventually Hosts, diagnosing IMP failures, and coordinating IMP software releases. The NOC also published network  traffic statistics to ARPA and via the RFC series to the NWG.  The NMC was interested in measuring the internal performance of the network under both real and artificial load, looking at things like queue lengths, packet lifetimes, maximum achievable throughputs, end-to-end delays, and so on.  In some sense they were responsible for telling ARPA whether the network BBN delivered met the specs of the RFQ.  They were also interested in comparing the actually performance to the predictions of queuing theory.  They made extensive use of traffic generation tools and statistics reporting tools built into the IMPs by BBN as "fake hosts", and collected  and analyzed the data on their Sigma-7.  Their experiments were often disruptive of network performance as seen by other Hosts, and therefore their experiment times were scheduled and controlled by the NOC.  This led to a certain degree of tension between the personnel at the NOC and the NMC, but nothing that couldn't be handled.  I believe the first publication of the NMC's work was a paper by Gerald Cole (one of Kleinrock's students) titled "Performance Measurements on the ARPA Computer Network" in the Proceedings of the ACM second symposium on Problems in the optimizations of data communications systems January 1971.
Hope this helps,Alex





    On Monday, September 7, 2020, 2:55:50 PM EDT, the keyboard of geoff goodfellow via Internet-history <internet-history at elists.isoc.org> wrote:  
 
 would be curious to know WHAT did UCLA-NMC measure and HOW did it measure
it?

i.e. was there a "precursor" to some kind of SNMP capability that allowed
UCLA-NMC to peer inside IMPs or hosts?

or were their processes on (all?) the ARPANET hosts at the time that
collected data from their respective OS's that the UCLA-NMC machine would
then periodically poll to get the data, kind of like the Tenex RSSER job
processes did with respect to sharing load avg. data among themselves?

[btw, believe that unless Leonard Kleinrock <lk at cs.ucla.edu> is
on/subscriber to the IH list, any reply sent to IH would be black holed, as
the IH list is most likely configured (Joe can confirm) to only allow
submissions to it from its "subscribers"... ERGO, if Len does reply, even
if cc'ng the list, you'd need to then manually fwd it to the list for
everyone else to see.]

geoff


On Mon, Sep 7, 2020 at 8:36 AM Steve Crocker via Internet-history <
internet-history at elists.isoc.org> wrote:

> I don't think the NMC measured anything at the host level, but I could be
> wrong.  Vint and Len copied explicitly.
>
> Steve
>
>
> On Mon, Sep 7, 2020 at 2:30 PM Jack Haverty via Internet-history <
> internet-history at elists.isoc.org> wrote:
>
> > On 9/7/20 11:15 AM, Steve Crocker via Internet-history wrote:
> > > It would be interesting to know the
> > > transfer rate between MTIPs.
> > This is the kind of thing that I'd expect the ARPANET NMC (Network
> > Measurement Center at UCLA, as opposed to NOC - Network Operations
> > Center at BBN) might have been measuring.  I don't remember ever seeing
> > any results or raw data from Measurements, and don't know much about
> > that part of the History.  Were results published somewhere and did
> > such reports survive?
> > /Jack
> >
> > --
> > Internet-history mailing list
> > Internet-history at elists.isoc.org
> > https://elists.isoc.org/mailman/listinfo/internet-history
> >
> --
> Internet-history mailing list
> Internet-history at elists.isoc.org
> https://elists.isoc.org/mailman/listinfo/internet-history
>
>

-- 
Geoff.Goodfellow at iconia.com
living as The Truth is True
-- 
Internet-history mailing list
Internet-history at elists.isoc.org
https://elists.isoc.org/mailman/listinfo/internet-history