Recent content by Fint

if one client works, but another doesn't, I would assume the issue is the non-working client.

md5 isn't encryption, it is a hash, and hashes generally can't be reversed.

F5 is pretty much the top name in load balancing. My company has several sets of BigIP LTM units, doing load balancing for a fairly large ($500m/yr) ecomm company. They are VERY stable, and fairly easy to use/setup.

Like this? http://www.networkworld.com/community/blog/cisco-usb-console-ports

Also, perhaps you should investigate being able to scale outwards... a design of 10 firewalls that each do 10Gb/sec is much easier to scale than 1 that does 100.

We've been using the Cisco twinax cables for nearly all of our 10G; works great.

in BGP, active is bad; it means it is actively trying to establish a connection, but can't. Generally speaking, eBGP (external BGP) requires the other side to be one-hop away, as a security mechanism; this could definitely cause you problems (unless the BGP config for the peer specifies it is...

On your PC, start three cmd prompts, and in each, ping a different target (your default gateway on your LAN, your 2nd hop, your 3rd hop) and see where the latency/loss is coming from. If you have zero loss to your gateway, and zero loss to your 2nd hop, but loss to the 3rd hop, you know the...

For what it is worth, I have GNS3 0.7.4 running perfectly on a Win7 Ultimate 64-bit, with zero problems. I usually have between 4 to 8 routers running at a time. I believe it is straight out of the box install, no modifications needed. I only use c3640-ik9s-mz.123-26.bin for all my routers.

TCP/443 is for the AnyConnect (SSL-based) VPN; the more traditional VPN is IPSec-based, which doesn't use TCP/443. I would think it would work, assuming whatever device is in front of the ASA isn't munging the packets.

The way DHCP works is the client boots up and sends a packet to 255.255.255.255 (the broadcast address) asking for any DHCP servers to give it an IP. The DHCP server then replies to the client's MAC address (they don't have a valid address yet), so they have to be in the same subnet (usually)...

I know this is trivial, but have you made sure you're running the latest NIC drivers on both NICs? This sounds very much like a bug in something.

iperf doesn't use your hdd; mostly network with a little bit of CPU. what's the latency between devices? You can try increasing your TCP window size; when I do iperf test I usually bump it up to 64k or so (iperf -w 64k) C:\>iperf -w 65k -c 192.168.220.87...

I don't know if this is even possible; a traditional router doesn't keep track of sessions, as it isn't stateful.

what version of AnyConnect? How are you installing it (via a push from a website, or directly from an exe)