[Info-vax] strange LAT problem

[Bob Koehler]

In article <puevab$ril$1 at Iltempo.Update.UU.SE>, Johnny Billquist <bqt at softjar.se> writes:
> On 2018-12-03 14:19, Bob Koehler wrote:
>> In article <pu0os2$tgf$1 at gioia.aioe.org>, helbig at asclothestro.multivax.de (Phillip Helbig (undress to reply)) writes:
>>> In article <pu0igt$54o$1 at Iltempo.Update.UU.SE>, Johnny Billquist
>>> <bqt at softjar.se> writes:
>>>
>>>> On 2018-11-28 21:52, Phillip Helbig (undress to reply) wrote:
>>>>> In article <ptls10$hsa$2 at Iltempo.Update.UU.SE>, Johnny Billquist
>>>>> <bqt at softjar.se> writes:
>>>>>
>>>>>> Any WiFi in this picture?
>>>>>
>>>>> No.  There is WiFi on the same LAN, but nothing to do with VMS.
>>>>
>>>> Meaning you use LAT between VMS hosts, and all the VMS hosts are sitting
>>>> on the same ethernet segment with nothing else in between?
>>>
>>> Right; just a switch.
>> 
>>     LAT is extreemly sensitive to timing, and it's possible that the
>>     switch does not meet the needs of LAT.  We've seen network bridges
>>     that required special firmware to hande LAT.  (Not routers, LAT is
>>     not a routable protocol, one can expect it to stop at a router.)
> 
> I would sortof disagree.
> 
> Yes, LAT is local network only. And yes, you can have weird network gear 
> that can cause issues.
> 
> But I am in fact bridging ethernet halfway around the world, including 
> LAT, and it actually works fine.

   I don't see the smoking gun, either.  But I've seen ethernet bridges
   deliver LAT packets out of order.

   That has nothing to do with latency.