From mboxrd@z Thu Jan 1 00:00:00 1970 From: Alex Kiselev Subject: Re: [PATCH v5 1/2] librte_lpm: Improve performance of the delete and add functions Date: Mon, 16 Jul 2018 18:36:25 +0300 Message-ID: <192110016.20180716183625@therouter.net> References: <5b4c51d3.1c69fb81.7a4de.519aSMTPIN_ADDED_MISSING@mx.google.com> <20180716081732.05302ae4@xeon-e3> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: "dev@dpdk.org" , Bruce Richardson To: Stephen Hemminger Return-path: Received: from relay-out4.mail.masterhost.ru (relay-out4.mail.masterhost.ru [83.222.12.14]) by dpdk.org (Postfix) with ESMTP id 11D5123C for ; Mon, 16 Jul 2018 17:36:41 +0200 (CEST) In-Reply-To: <20180716081732.05302ae4@xeon-e3> List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org Sender: "dev" > On Mon, 16 Jul 2018 11:05:27 +0300 > Alex Kiselev wrote: >> librte_lpm: Improve lpm6 performance >> Rework the lpm6 rule subsystem and replace >> current rules algorithm complexity O(n) >> with hashtables which allow dealing with >> large (50k) rule sets. >> Signed-off-by: Alex Kiselev > Internet routers can have 1M rule sets. > The cost of a hash table maybe too high then? If I am not mistaken cuckoo hash algorithm on which the rte_hash is based is just a plain array logically divided into buckets. So, I am not introducing any memory overhead in comparison with current rule storage implementation that uses a plain array. -- Alex