From mboxrd@z Thu Jan 1 00:00:00 1970 From: kernel neophyte Subject: Re: rocksdb with jemalloc Date: Tue, 12 May 2015 14:09:38 -0700 Message-ID: References: <55516BB9.6000709@redhat.com> <6F3FA899187F0043BA1827A69DA2F7CC021EF61E@shsmsx102.ccr.corp.intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Return-path: Received: from mail-pa0-f51.google.com ([209.85.220.51]:35330 "EHLO mail-pa0-f51.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752020AbbELVJi (ORCPT ); Tue, 12 May 2015 17:09:38 -0400 Received: by pabtp1 with SMTP id tp1so26417355pab.2 for ; Tue, 12 May 2015 14:09:38 -0700 (PDT) In-Reply-To: <6F3FA899187F0043BA1827A69DA2F7CC021EF61E@shsmsx102.ccr.corp.intel.com> Sender: ceph-devel-owner@vger.kernel.org List-ID: To: "Chen, Xiaoxi" Cc: Mark Nelson , "ceph-devel@vger.kernel.org" On Tue, May 12, 2015 at 1:11 AM, Chen, Xiaoxi wrote: > Hi Neo, > Did you hit the error when doing configure? The error message looks like > configure: error: --with-jemalloc called with --with-librocksdb_static, turn off > --with-librocksdb-static or --with-jemalloc Yes, this is what I am talking about. > This is because RocksDB use tcmalloc so we cannot build it with jemalloc. Thats not true ? RocksDB can be made to build with jemalloc. I will send a PR soon. > But if you manually clone RocksDB from facebook , compile it and install, then use ./configure --with-librocksdb --with-jemalloc, seems we can pass the build. > I tried this, build succeeds, however newstore keeps crashing with this approach, I havent got around to why it crashes. -Neo > Xiaoxi > >> -----Original Message----- >> From: ceph-devel-owner@vger.kernel.org [mailto:ceph-devel- >> owner@vger.kernel.org] On Behalf Of Mark Nelson >> Sent: Tuesday, May 12, 2015 10:56 AM >> To: kernel neophyte; ceph-devel@vger.kernel.org >> Subject: Re: rocksdb with jemalloc >> >> On 05/11/2015 08:41 PM, kernel neophyte wrote: >> > Hi, >> > >> > >> > I am trying to use newstore + rocksdb + jemalloc. Currently rocksdb in >> > ceph compiles only with tcmalloc and not with jemalloc. >> > >> > Is there any reason for this limitation ? >> >> Heya, I'm guessing it's not by design, simply no one's tried doing it yet. What >> are you running into? >> >> > >> > -Neo >> > -- >> > To unsubscribe from this list: send the line "unsubscribe ceph-devel" >> > in the body of a message to majordomo@vger.kernel.org More >> majordomo >> > info at http://vger.kernel.org/majordomo-info.html >> > >> -- >> To unsubscribe from this list: send the line "unsubscribe ceph-devel" in the >> body of a message to majordomo@vger.kernel.org More majordomo info at >> http://vger.kernel.org/majordomo-info.html