On Thursday 23 March 2006 09:31, David S. Miller wrote: > The key point is to keep the per-socket limits far enough away from > the global pool limits such that it is not easy for a single entity > to maliciously put the allocator into conservative mode and penalize > the legitimate users.
It's probably a lost fight already because many services have legitimate needs for many sockets and any number for max sockets you pick for max socket buffer = main memory for tcp / max sockets will not fit some workloads. But normally only a small number of sockets go anywhere near the max socket buffer. The dynamic throttling of send windows that is already implemented looks quite good to handle it, but perhaps it just needs some tuning to be a bit more aggressive. -Andi - To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
