Mark put forth on 6/21/2010 11:13 PM: > On Mon, Jun 21, 2010 at 9:03 PM, Stan Hoeppner <s...@hardwarefreak.com>wrote:
>> Thus, with the same router, I could take a few different *nix OS flavors >> and >> perl versions, blowing up the router with some, and not denting it with >> others. >> >> It's all about the packet load you push through the router. It's >> absolutely >> normal for setups that "seem" the same to nuke the router, because once you >> peek under the hood, they aren't really behaving the same at all. >> >> Take a peek under the hood. :) >> > > Interesting, so is the router to blame or the OS? The answer isn't a simple either/or. You can't simply lay blame either. It's a balancing act. If you're of the opinion that any consumer router should be able to take anything you throw at it, then the router is to blame. But we all know that "you get what you pay for". In that sense, it's not the router's fault but the customer's for buying "cheap" or "less than capable". Sure, the consumer didn't know it at the time of the router purchase, but hay, that's life. That's why Walmart sells a $69 lawnmower and a $269 lawnmower. They both should be able to mow any amount of grass on any terrain _forever_ without breaking right? (laughs) _WRONG_. The $69 mower will last an owner of a large yard for a season, maybe two. Then it will fail. One personality type will draw the mower back to Walmart and scream and shout at the customer service people making all kinds of demands. The other personality type will realize he bought a cheap fucking mower which failed after two seasons, and he'll go back and buy the $269 mower which may likely last 10 seasons. Over 99% of all consumer broadband users have a "$69 mower" that their ISP gave them free of charge during service connection. Somewhere between 1% and 10% of these users really hammer their free "$69 mower", then complain when it fails to perform the way they expect. > Because you're fixing the > problem by the OS, not changing the router. No, I'm actually fixing the problem by changing the application to work around the limitations of the routers. When mentioned some combos break the router and others do not, picking one that doesn't isn't a solution. The next aptitude safe-upgrade may cause one's apps to start breaking the router. > Short of people buying beefy > commercial grade routers for home usage torrent downloading, what's the > solution? _IF_ indeed Debian/Ubuntu users are nuking their routers with torrent traffic, then the solution is the same as the rDNS tool solution I use to keep from UDP flooding my router: You modify the application, or tweak its settings (if such settings are tweakable), to keep it from melting the router with its traffic pattern/load. Simple. _IF_ it is unacceptable to such users to have to "slow down" their torrents, and thus they aren't willing to change the packet behavior of their app, then they simply have to pony up and buy a better router that doesn't melt under the load. If this is truly a problem in the wild with torrent users, there will be thousands of forum and list posts on the net containing lists of models of wired and wireless routers that have been verified as good choices for torrent users. I'm not a torrent user so I have no clue what goes on in this world. However, just like everywhere else, if there is a widespread problem in a community, there will exists plenty of information online about said issue. In the case of the original OP, he doesn't own or control his router, so this is not an option for him. His only option is to find an OS/application combo that doesn't break things, or tweak one that is breaking things until it no longer does so. -- Stan -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/4c204066.90...@hardwarefreak.com