RE: Wired vs wireless - comm trace needed

Here's the latest from Zebra on this:

Hi Jeff,

Sorry for the delay in response.

Here's the situation - we've found that there are SNMP packets in the data
stream being directed to the printer that are not exactly to standard. To be
more specific -

The AS/400 sends a non-standard SNMP getnext request. The AS/400 getnext
request contains 10 variable-bindings. The AS/400 only seems to be
interested in two of those varbinds, the remaining eight are "null"
varbinds. What I mean by "null" is that each one is simply two "zero" bytes
tacked onto the end of the SNMP packet...a total of 16 nulls get tacked on
the end of the packet.

So, knowing all this, we altered our SNMP test app to work like the AS/400
does - this allowed us to reproduce the issue. The theory is that the
printer slows down because it tries to perform a getnext on each of the null
varbinds, and this, combined some network errors and retries we saw in your
packets, occupies the printer as it's network stack tries to make sense of
all this.

It's important to note that the same type of packet structure is being
directed at the wired print server, but obviously the same issue is not
happening. This is, in our opinion, for two reasons. The first is that the
two devices (wired and wireless) use different networking stacks. The second
is that the wired print server runs on its own processor unlike the wireless
print server that runs in the same processor environment that the printer
firmware operates within.

Having said that, the slowdown should not be happening. What we've done is
to alter our wireless network stack to deal with this scenario more
gracefully. That new version is in test now. I believe it will be in a
sufficiently validated to send it over to you for a test by next week.

We've actually not seen this type of structure in AS/400 SNMP traffic before
- it might be worth asking IBM about it if you have a way of doing so. In
any event, we'll still make the altered firmware version

I hope this helps explain what is going on. Please let me know if you have
any questions.

-Leo

I opened yet another PMR on this with IBM and passed on Leo's email to IBM.