Linksys WRT54G slow loading webpages

sounds like spyware or so called browser helpers. if the files are getting through fast your router is not the issue.

Reply to
frank
Loading thread data ...

Is it slow at loading individual elements or is it slow at completing the entire page?

If the latter, web pages these days get files from many different locations and it may be that your system or its DNS servers are no longer caching IP addresses (there was a major issue with certain versions and configurations of BIND a few weeks ago). Thus, there may be a delay as the download of each element starts because of DNS lookups. The solution may be to switch to a browser that can be configured to download many files simultaneously.

If the former, some ISPs limit the bandwidth for data that travels over certain ports (I think that the WRT54G also has a similar feature, but it needs to be turned on and it is not the default). Thus, it may be that your ISP is limiting the amount of traffic over port 80.

Finally, are you absolutely sure that it's not spyware?

-Yves

"???"

Reply to
Yves Konigshofer

"???"

Reply to
Floyd L. Davidson

"???"

Reply to
Floyd L. Davidson

Has anyone else had this problem? For the past few weeks now I've notice that my router is very slow loading webpages, ibut f I download a file the speed is great. Need help. The router is a Linksys WRT54G ver 2. with the latest firmware.

Reply to
???

Trust me, it's not spyware.

Reply to
???

"???"

Reply to
Floyd L. Davidson

"???"

Reply to
Floyd L. Davidson

I'm sure it's not spyware, the problem been happening on all of my pc's on the network. Plus the laptop I use just had a fresh install of the OS.

Sometimes the page loads slow and sometimes it doesn't load at all. It even loads slow sometimes when I log in to the router setup page. And it doesn't matter if I use wireless or wired connection.

Reply to
???

It doesn't really mean that anything is screwed up though. The point is that each node is *supposed* to be able to accurately fragment any packet, and at the end it all gets properly reconstructed. The problem with some xDSL modems is they

*aren't* able to fragment properly. Instead the PPPoE module gets a 1500 byte packet and wants to tunnel it to the router, which of course requires an added header for the tunnel protocol, and makes the packet larger than 1500. It should be slit into two packets, but the modem doesn't do it right.

Note that 1500 is the standard MTU for use on a 10Mbps Ethernet... however, at any speed slower than that it is better to have a smaller MTU. In fact, the specification for forwarding says the receive buffer only has to be somthing 576 bytes, and as a result it is quite possible for any packet larger than that to be lost! That's why you'll see a lot of Internet software that defaults to 576 for the MTU. DEC used to use 1004 as default. And anyone with a v.34 modem should set their PPP to used an MTU down in the 296 range if they do any kind of interactive multitasking on the link.

Conversly at speeds greater than 10 Mbps a larger MTU is reasonable too. The lo interface on Linux uses 16436.

Reply to
Floyd L. Davidson

Between nodes, which certainly includes routers. Any interface that actually looks at the IP addressing, will be negotiating the MTU. That basically amounts to any interface which has an assigned IP address itself. Hence an Ethernet hub (which ignores all addressing) or an Ethernet switch (which looks at MAC addresses) will not fragment IP packets and will not have an associated MTU.

In some cases that would be exactly it. If your PC's interface is your last IP interface... which it would be if you, for example, use a dial up modem and run PPP on the PC. It won't be if you have a separate firewall, and an xDSL modem qualifies for that. The modem has it's own IP address, and looks at the IP addressing in packets sent in either direction.

Actually, with Point-to-Point protocols (any kind of PPP), it is guaranteed that the two ends have negotiated an MTU when the link was initialized.

Fairly typical. 576 is the minimum buffer size required by the original protocols for IP nodes. Any node was supposed to be able to handle a datagram of at least that size. The idea was a

512 byte data payload, and up to 64 bytes of protocol headers. (An IP header is a minimum of 20 bytes, and could be as much as 60.)
Reply to
Floyd L. Davidson

On Thu, 12 May 2005 03:17:39 GMT, "???"

Reply to
Jeff Liebermann

thanks for the tip

Reply to
???

Reply to
Jeff Liebermann

No problem

Reply to
???

On Thu, 12 May 2005 04:37:44 GMT, "???"

Reply to
Jeff Liebermann

Since I downgraded the firmware, I'm getting fast page loads again. I think it was just a bad firmware. I hope that's the problem.

Reply to
???

Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.