Ping has nothing to do with FPS. If you're having FPS problems, that's a hardware issue with your computer - probably video card.
IME, 200 ping is higher than desirable, but not worthy of a complaint. 400 or 800 or 1200 is though, and that's what I'm getting....
To trace your route: First, run GW, then alt+tab back to windows and open a command prompt. Type "netstat -na" followed by enter. Now alt+tab back to GW and log out. Under the foreign address column, you should see 3 entries that read X.X.X.X:6112. These are the servers you are connected to in order to run GW. Pick one. Now type "tracert X" where X is the IP of the server you chose. This will report the route and the ping at each router along the way.
Reading the report:
- (Zeroth of all, "downstream" means "closer to you," and "upstream" means "farther away from you.")
- First of all, understand that answering ping requests is (almost) always given lower priority than passing along data (including ping requests addressed to someone else), so a moderately congested (or worse) router will pass along your pings addressed to upstream routers faster than it will answer your pings addressed to it. That's why it's possible for an upstream machine to have a lower ping than a downstream machine or an upstream machine to respond after a downstream machine timed out in answering your request.
- A jump in ping at a downstream router followed by a lower ping at an upstream router usually means nothing but congestion at the downstream router. In the worst case it means you'll need to use pathping to check for packet loss if the congestion is very bad.
- Conversely, a jump in ping at an upstream router followed by pings that are as high or higher at every upstream router means that there is real slowdown there.
- Real slowdown is normal, and should be expected, when it correlates with geography. (Example: A hop from Chicago to Dallas is going to have real slowdown, since even light moving down fiberoptic cable only goes so fast.)
- Conversely, real slowdown where geography provides no explanation and/or real slowdown to a speed that is simply unacceptable are strong indicators of the location of a problem.
- Timeouts can mean one of four things:
- Case 1: The router is totally overloaded or malfunctioning or the physical connection is broken. If this is the case, you won't see anything past this point and will be totally unable to connect to anything upstream of it in any way. Obviously, this is bad. Hard to distinguish from case 2, below.
- Case 2: The router is heavily congested enough that your ping request (low priority) never gets answered, but (some) traffic is still getting forwarded, but so slowly that ping requests to upstream routers don't make it roundtrip before the timeout deadline. Although not as bad as no connection at all, this is about as bad as connection problems get. Hard to distinguish from case 1, above.
- Case 3: The router is configured not to repsond to ping requests, so it isn't responding to you. This is not a problem at all. You will usually see responses from upstream routers following a timeout if this is the case. This can be hard to distinguish from case 4, below.
- Case 4: The router is heavily congested enough that your ping request (low priority) never gets answered, but (some) traffic is still getting forwarded, albeit slowly. This is nearly as bad as case 2, above. You will usually see responses from upstream routers following a timeout if this is the case. This can be hard to distinguish from case 3, above. Probably the best way to distinguish them is to look at the ping on the next router upstream - if there's a jump, then probably case 4; no jump, then probably case 3.
If tracert isn't enough to pin down the problem, run "pathping X" where X is the IP of the GW server you just tracert-ed. Expect it to take a good 5 minutes. Pathping will return the same info as tracert, plus raw figures for packet loss percentages of packets sent to each router, plus a microsoft algorithm's (somewhat dubious) assessment of which routers or cables caused what percentage of the packet loss.
- Routers with both unexplained jumps in ping and any packet loss at all are prime candidates for being the source of a problem.
- Any packet loss at all is suspect, even if there is no ping problem. Packet loss over 1% or 2% is especially suspect and probably indicates a problem.
- Like a timeout, packet loss of 100% (raw, not microsoft's attribution) can mean several things:
- No connection at all
- Bad enough congestion that no ping requests are getting answered (and some or all through traffic may or may not be getting dropped too)
- The router is configured not to reply to ping requests
(Hmmm.. maybe I should copy this out and call it a guide...)
Reply With Quote
Originally Posted by Matrix1160
It can take a few hours if it's really bad. I believe all of the third party defraggers have schedulers that'll let you setup a regular defrag run when you're not around. The built-in would require a bit more work to get it to work with the scheduler (no point-n-click scheduling).
