Recent content by jroysdon

Speaking of dust problems on a PIX. About a year ago our PIX 520 was making some noises that sounded to me like a fan dying. The PIX was sealed, so I didn't want to just slice the warranty sticker, so I contacted Cisco, and they wanted like $1k to cross ship us a replacement, or around $750...

I believe he's on the same LAN. The problem is that STP is putting the ports on the 2924 in blocking mode when the PC's first power up, and given that they probably power up very fast, the port doesn't go into forwarding mode when the DHCP request tries to go out, and thus the problem. Two...

Be more specific. A broadcast is local to a subnet. The DMZ would be another subnet, and therefore wouldn't be part of the broadcast. The reason I say be more specific is the only exception I can think of would be for a directed broadcast. The default for new router IOS is 'no ip...

I can't speak in absolutes for all Cisco platforms, but my 1605R & 804 both use ntp's standard udp/123 port. I run a very tight config (IOS Firewall and an ACL that blocks everything except what I allow or the IOS Firewall opens up on the way out), and I know the NTP wouldn't work until I added...

Post your config and we might be able to help more (remove passwords, SPIDs, etc.).

Try removing ppp multilink and only dial out with one channel (remove the second SPID)and see if you get 56 or 64kb as you should. If so, I'm guessing the channels aren't bonding properly, and it's doing some sort of round-robining out both devices and half the packets are getting dropped/being...

Nuke the ACLs on the console port and see if it doesn't just magically get fixed (you said you still have telnet access, right?). If you don't have the ACL defined, I'm guessing it's doing an implicit deny any. I can't see why you'd even want an ACL on a console port. I can understand on a...

First, 192.168.0.0 will route on a LAN the same as any other non-class D/E or loopback address. It may not route through your ISP across the internet, but that's only because the ISP is chosing to drop those packets (as the RFC's say those certain private address ranges shouldn't be routed)...

Bandwidth has no actual effect on the bandwidth. It's used by routing protocols to determine metrics, etc. The main lines needed is the clockrate and the following line on the DCE side: dce terminal-timing enable

I assume the IBM is remote, therefore the need for ISDN. Get a router than can support token ring interfaces (some 2500s) and connect an external ISDN unit off of a serial interface. If you want to spend a bit more cash, get a 2600 and add a token ring interface and ISDN BRI WIC.

I don't see any idle-timeout configured. The syntax is: dialer idle-timeout 240 (4 minutes, as it is specified in seconds)

A minor correction: the biggest difference between the 1600 and 1700 line is the CPU performance. Also, the 1720 comes with no interfaces beyond the single ethernet, which the 160x comes with two interfaces (my preference is the 1605R with a T1 WIC as it comes with two ethernet ports and one...

You'll need a WINS server at the central location or use LMHOSTS which are stored: Win9x: c:\windows\lmhosts NT/W2k: c:\winnt\system32\drivers\etc\lmhosts You'll need a domain master at the remote location pointing at a WINS server if you want Browsing to work. Even a Win95 box can act as...

Cisco TAC and UniverCD: http://cisco.com/univercd/cc/td/doc/product/software/ios120/12cgcr/secur_c/scprt3/scfirewl.htm#8179

I know this is an old post, but I'd feel a lot safer if all of my boxes were inside the firewall to some degree. If you're really paranoid about someone hacking the Webaccess box, put it on a DMZ leg, but don't put it out in the open as it still has to talk with the GW server, which means it...