What a joke... assigned an invalid IP

[who.is.michael]

Assumptions, assumptions…

The only thing I can think of is a faulty switch somewhere (3Com comes to mind) but it seems recently there have been complaints from joburg users, so either they are starting to use the same switches as they use in cape town or perhaps it is poor integration with the UUNet / Verizon network, but this is all speculation.

Don't think iBurst knows, as long as nobody complaints continually I don’t think they will do something about it, but yeah I would also like to know what the cause is - if they ever do find out. (I've given up on hassling them, I only call them now if the BW is down to 100 kbps and usually speak with Ed, so now since he's leaving I'm not sure what to do.)

 

[w1z4rd]

Maybe it has something to do with the links between towers? They use lasers to connect the network together and from what I understand (which is very little) this can cause bottlenecks.

 

[MrH]

Even if they find the problem I doubt that they will tell us since thats admitting a bad service.

Going to get some sleep now as I get up at 7:00am

Seems you also burn the midnight oil.

 

[who.is.michael]

Sleep Well MrH - Interesting theory w1z4rd, how much do you know about their tower infrastructure? Do you know what type of antennas they use? Never seen them elsewhere and cant find the supplier?

 

[w1z4rd]

Sleep Well MrH - Interesting theory w1z4rd, how much do you know about their tower infrastructure? Do you know what type of antennas they use? Never seen them elsewhere and cant find the supplier?

They bought the Iburst technology from a group of Japs... so look in that area. If you look at earlier Iburst posts, you will see the "Japanese" comments. I would imagine their towers are very specific to the Iburst technology.

I have seen the laser link boxes they put up, and have seen the laser boxes at their offices. I dont know enough on the technology, but people I have spoken to me who's opinion I listen to very carefully recon thats where a major bottleneck occurs.

 

[MrH]

w1z4rd, what you say is possible except the IPs I'm talking about are in the same condition whether it's during the day or after midnight when traffic is low.

 

[Next Thing]

I spoke to my employer's IT staff and they agree with me that I am right about this:

They are wrong ...

Depending on your network masks and adress ranged it is perfectly fine to have a ip address that ends in .255

 

[JTech]

Depending on your network masks and adress ranged it is perfectly fine to have a ip address that ends in .255

Yes indeedy, if you use the line for access to the internet, and not in "reverse" i.e. to host servers. I am not going to repeat myself as to why .255 is invalid for me. The line works for God's sake, client software falls over, THAT"S MY PROBLEM I don't know how to explain it any easier.

No-one can seem to understand me so I've asked the mods to delete this thread. Funny how the lot on prophecy can understand what I mean, but you lot are unable to.

 

[WHATEVER]

hehehehe ... getting a little tense in here!!!

You CAN use .255 ... but best practice say you should not. So apps are written to ... not use .255.

I think that is wrong ... but hay ... who am I

 

[Next Thing]

Yes indeedy, if you use the line for access to the internet, and not in "reverse" i.e. to host servers. I am not going to repeat myself as to why .255 is invalid for me. The line works for God's sake, client software falls over, THAT"S MY PROBLEM I don't know how to explain it any easier.

No-one can seem to understand me so I've asked the mods to delete this thread. Funny how the lot on prophecy can understand what I mean, but you lot are unable to.

I understand you perfectly ...

I have yet to see a program that doesnt work with funny ip range allocations. The only time I have seen problems is when the Broadcast and Network Address ( ie the .255 and .0 in a class c based range ) are incorrect. The routing of traffic is not done by the application but by the tcp stack itself.