Author: Steve Coladonato
Date: 06:05:11 03/09/00
Go up one level in this thread
On March 09, 2000 at 08:46:20, Robert Hyatt wrote: >On March 09, 2000 at 07:59:16, Steve Coladonato wrote: > >>On March 08, 2000 at 14:59:31, Robert Hyatt wrote: >> >>>On March 08, 2000 at 12:37:44, Steve Coladonato wrote: >>> >>>>On March 08, 2000 at 12:10:01, James Robertson wrote: >>>> >>>>>On March 08, 2000 at 11:21:38, Steve Coladonato wrote: >>>>> >>>>>>Game was adjourned after move 10 awaiting the arbiter's decision. >>>>>> >>>>>>It looks like matches on the internet are not ready for prime time live although >>>>>>I don't see the same type of technical issues on FICS. >>>>> >>>>>When will it start again? >>>>> >>>>>James >>>> >>>>I believe it's up to the arbiter. He is supposedly going to determine how the >>>>rest of the match will continue. >>>> >>>>The Linares tournament doesn't seem to have connection problems (at least from a >>>>viewer standpoint). I minimize the game window and periodically recheck it to >>>>see how things are progressing. >>> >>> >>>The Linares web site has actually been bad. It has gone down several times (it >>>can't be pinged or anything). >> >>In general, are long term connections to an ISP a major problem with the current >>state of the technology? I get the impression that FICS works because the >>connection time to complete a game is generally less than 30 min total. >> >>The Internet tournaments that have recently started are using more conventional >>time controls and the connection time to complete a game is measured in hours, >>not minutes. I think, the problem should be worked out, rather than generate >>the frustation that is happening now. There were a few published problems with >>games in the Club Kasparov tournament and now the same problem in the Xie Jun >>match. The Linares problems are the same, but only the viewer is affected, not >>the players. >> >>Steve > > >Not that I know of. IE crafty/scrappy are regularly connected for weeks at a >time with no problems... > >These problems are caused by at least a couple of things: > >(1) the host machine isn't up to the load it actually gets, due to poor planning >or exceptional demand. > What machines/os are used as hosts at most of these sites? (Intel/NT, Intel/Linux, HP/HP-UX, Alpha/NT...) No one probably knows for sure but an educated guess would be appreciated. I'm sure these events are generating interest in the chess community and world wide there would have to be quite a few simultaneous connections to these sites during the events. This would make load testing almost a necessity before hosting events like these and you are most likely correct that the testing is not being done. >(2) the host is behind something that serves as a serious bottleneck to IP >traffic, causing packets to get dropped. > This is beyond me technically but it is certainly a valid point. >(3) the host software is lousy, causing the server to crash under a strain. > This, however, is more common than one would think. >There are 'tools' that can simulate heavy loads, but they obviously don't get >used very much. :)
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.