<html>
<head>
<meta http-equiv="content-type" content="text/html; charset=utf-8">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<p>Thanks for your interest, And your participation last Tues.<br>
</p>
<p>I am breaking this little report down into sections. This is the
first.</p>
<p>You have to know that the interest in the topic of discussion is
much higher than anyone might have predicted.</p>
<p><u>What happened During the net last Tues..</u></p>
<p>Let me say that this was originally configured for 60-70 nodes to
connect. Should not have been<u> any</u> issues.<br>
</p>
<p>That is what I was expecting with the unknown of personal/Pi
nodes that are quite numerous now if you have not looked.</p>
<p>Being a repeater or a personal Pi, ...it's a stream of the same
size .<br>
</p>
<p>A guideline for connecting was published. Those that connected in
advance adhered to it.</p>
<p>I made one balancing move in the beginning before the net and
nobody even knew it happened.</p>
<p>This was to keep nodes on par with their potential bandwidth. I
had some options to balance even further.<br>
</p>
<p>At some point near the beginning of the net, about the time
Steve/N4IRS was having issues with the digital input and time-out
on his own node, ...well...some took that as a node drop-out and
began to switch connections from the planed path.</p>
<p>At this same time, we had a influx of connections (while net was
ongoing) LATE COMERS ! <br>
</p>
<p>Expected some, but there were many. And most were not adhering
to the connection plans.</p>
<p>Within that part of the mess, a node had been connected up who
came with 50+ nodes in a private network. I could not identify
this node for the duration of the net. No way to see who is
spewing large chucks of iax data on our network. It's not going to
show in your allmon. This unidentified network normally runs with
all their elemetry "OFF" (found out later)<br>
</p>
<p>Most of you know how telemetry works in the iax2 and know your
connected brings that data to your network. Same as they will
receive yours. But this one node had telemetry data beyond belief
and in loop. They don't know it, it's turned off on their
systems. With most of our nodes are using telemetry, it became a
additional overload load for Allison just trying to speak it.</p>
<p>And the telemetry list stacked-up beyond comprehension. And very
quickly.</p>
<p>With all of the cpu activity and iax activity, I could not get a
command in edge-wise. At this point I don't know where the iax
data is coming from or that it is a major factor in the eventual
overload of 29999. I had figured that much of it was coming from
one node because it was the same stuff over and over. It's a funny
story how I discovered the who but I'll save that for some other
dicussion.<br>
</p>
<p>I have some very strong feelings about this -<br>
</p>
<p>'running a large network without any telemetry and connecting to
other large networks"you better be sure your network is clean, but
I'll keep that under control for now.</p>
<p>Everyone should realize that when you connect to others,</p>
<p>you bring any problems your system has to the network you connect
to. <br>
</p>
<p>Even if it is not noticeable to you when connected to anything.<br>
</p>
<p>I take full responsibility for this failure. But I want to share
some of the blame on some others.</p>
<p>Those that will not get with the plan,<u> especially when late.</u>
And those that will not stick to the plan making matters worse.</p>
<p>Contingency's were in place but you can not make changes when
everyone is moving to who knows where.</p>
<p>And that iax data stream kept me from changing anything anyway.<br>
</p>
<p>I needed some direct communication with Mike and Steve to make
them aware of things during the net process and pause them until
we made adjustments. But this came about very suddenly and
snowballed.</p>
.<br>
<p>Normally we would account for some of these late connections but
there was a large number of them.</p>
<p>--<br>
</p>
<p>Most of you can match that from what you seen and heard to
understand this better.</p>
<p>But on my next note to group, I will post my plans for the next
net. They are very strong. I will post by Saturday/Sun.<br>
</p>
<p><u>Then some new thoughts for discussion going forward in this
new age of personal connectable network systems</u>.</p>
<p>We have twice as many nodes now as just 3-4 years ago. It's still
growing at that rate++. The older days when repeater systems were
the dominate bunch are over. Repeater nodes were pretty much
predictable. Personal nodes are not.<br>
</p>
<p><u>We need to have this discussion before we grow to big for it
and things get to messy to keep the good perception we recieve
from others at this point..</u></p>
<p>Watch for my next post on this issue and you will see clearly how
I am going to deal with these issues going forward.</p>
<p>It may be just one example of many to follow of how to avoid
these problems in the future. And know that every one of these
issues will be fully addressed in the next net. Since we can no
longer judge the size of interest and participation with any real
degree of accuracy anymore.<br>
</p>
<p>Thanks for reading the whole thing ! I welcome responsible
comments at this time. But I would ask you to wait and see the
plan going forward. These nets that 'anyone produces' are going to
get larger and we need a plan.</p>
<p>It is my desire to involve as many going forward as we can.
Because others will need the practice of knowing how to properly
configure and run such a net structure that continues to grow. And
it is growing rapidly. The more that have a feel for this, the
better we all are. Myself included.<br>
</p>
<p>...mike/kb8jnm<br>
</p>
<p><br>
</p>
</body>
</html>