HP 4050Ns Drop Off LAN Overnight - 3Com Switch Problem??

[250LWB]

I have a LAN that was recently changed over from 3Com hubs to 3Com SS II 1000 and 3000 switches. All has been well until a few days ago.

I have 9 - HP 4050N LaserJet printers on the network. Over the last two nights 5-6 of them have been "dropping off" the LAN, that is, they no longer show up on AppleTalk, Network Supervisor of anywhere else. If I go to the switches they are connected through the link lights are out on the ports they connect to. I think we have experienced power outages on both evenings but I'm not positive about that - I have to check my UPS logs and see. I don't see why that would make a difference anyway.

If I remove the cable from the switch port and plug it back in I get a link and all is well once again.

The printers do go to "sleep" during periods of inactivity but they have been this way for nearly two years without a hitch.

I'm thinking that something in the switch may be causing this.

Please help - I don't want to have to walk around campus every morning to plug and unplug cables to make the LAN happy. . .

250 LWB

 

[jimbopalmer]

can you set the printers to be 10 meg and half duplex? if so, try that. the printer does not really talk back that much anyhow to need full duplex. I am betting after the power comes on they are confused about speed and duplex, and if you set it to the 'default' of 10/half they never get confused. I tried to remain child-like, all I acheived was childish.

 

[250LWB]

Thanks for the suggestion. I'll check the settings when I'm back on campus in the morning.

What I find odd is that the other printers on the LAN are just fine, yet they have the same periods of inactivity the ones dropping off do, too. It seems to be centered around two particular switches fomr what I see. In both data frames I have a stack so I was thinking of reassigning the printers to another switch in the stack to see what happens.

250LWB

 

[jimbopalmer]

before taking my first suggestion be sure both the switch and printer are at autonegotiate now, maybe they are confused case they CAN'T autonegotiate I tried to remain child-like, all I acheived was childish.

 

[cgv]

I have a similar problem with devices dropping off the 2nd in a stack of 2 3300XMs. Manual power cycle corrects the problem.

Moving most of the network traffic off of the failing switch did not seem to help at all. It appears like the switch stops sending traffic through the matrix cable.

Flashed up to 2.69 with no effect. In fact, it may be worse.

I know this is no help... just noting a similar issue.

 

[250LWB]

It has been suggested that HP printers don't like to play nice with DCHP and should have fixed IP addresses.

That could very well be, but this was the first occurrence of this sort I've had with the same configuration for over 2 years. I have yet to experience it again.

Worst case scenario I will assign fixed IPs to all the printers and be done with them.

Thanks, to all,

250LWB