what gives with this darn thing?

[Ron Irving]

We have a remote office, connected via T1. Earlier this week, the T1 went down. When it came back up, some of the remote office were able to login, some were not. The only way to get the affected ones a login, was to ping their ip address from here, (the home office.) Once we did that, everything worked. Even some of the printers had trouble until we pinged them. Any ideas?


Thanks in advance,

Ron Irving

[Richard Darling]

Ron, If you use DHCP did you have them release and renew their leases...just a thought.

[Sachin Patel]

Hi Ron,
May be cache issue.

Sachin

[Craig Gilmore · Accepted solution]

You have a network device problem. Are the devices on the non-system end of the T1 line connected via a router or a bridge?

The ping added a routing/arp entry for the device in the link allowing communication to occur. Depending on the type of communication across the T1 line will explain the network level of the communication problem.

[Wodisch]

Hello Ron,

are you talking about PCs running M$ Windows* where
your users encountered login problems?
Well it could be that your routers (on "their" end) are
configured to *NOT* pass through broadcasts.
But M$ login service works with IP-broadcasts...
So those PCs cannot find your PDC oder BDC (primary
or backup domain controllers), probably on your side of
the T1.
Your pings did enter the needed information into the
caches of those PCs then.

HTH,
Wodisch

[Jay Newman]

Certain network software packages will mark the default gateway for deletion if the connection to it goes down, and will not put it "up" again unless you restart the network or reboot.
You may need to issue commands specific to each O/S to re-establish the default gateway.
If only some of your systems cannot connect, is there a pattern such as being inside a particular subnet, or outside of the subnet of your host?
You can always use the route command specify the gateway to a particular address or range of addresses; if it fixes the problem, you know there is a missing or incorrect default gateway somewhere.

[Anthony deRito]

Not sure if you have considered this but this may be a routing issue. The best way to determine if there is a routing issue is to perform some traceroutes to a client who cannot connect. See which hop is not responding. Compare this to a good traceroute to a connected host. Run netstat -nr to check local routing table and make sure you have a route to your gateway. Also check the routing tables on your router that the T1 comes into in addition to those at each hop. With Cisco, its "show ip route"

Tony