From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mark Nelson Subject: Re: [ceph-users] rocksdb Seen today - replacement for leveldb? Date: Wed, 27 Nov 2013 08:54:20 -0600 Message-ID: <5296079C.3060101@inktank.com> References: <5295CF96.6040708@profihost.ag> <5295CFE2.8070603@profihost.ag> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit Return-path: Received: from mail-ie0-f177.google.com ([209.85.223.177]:52554 "EHLO mail-ie0-f177.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752136Ab3K0OyQ (ORCPT ); Wed, 27 Nov 2013 09:54:16 -0500 Received: by mail-ie0-f177.google.com with SMTP id tp5so11986397ieb.36 for ; Wed, 27 Nov 2013 06:54:15 -0800 (PST) In-Reply-To: Sender: ceph-devel-owner@vger.kernel.org List-ID: To: Haomai Wang Cc: Stefan Priebe - Profihost AG , "ceph-devel@vger.kernel.org" Very interesting guys! I'm also interested in sync reads/writes. That's where leveldb really seemed to fall over in the testing I did. On my laptop with an SSD I was getting 200 iops no matter if I was doing random or sequential sync writes (using a single thread) with stock leveldb, hyperdex, or basho's version. It would be interesting to try rocksdb. Nisha Talagala of Fusion IO also did a presentation at PDSW comparing their OpenNVM key-value layer to leveldb for sync reads/writes here (slide 24): http://www.pdsw.org/pdsw13/slides/talagala-Flash_datacenter-pdsw13.pdf Mark On 11/27/2013 05:52 AM, Haomai Wang wrote: > Yes, we have related bp for rocksdb backend. > wiki.ceph.com/index.php?title=01Planning/02Blueprints/Firefly/osd:_new_key%2F%2Fvalue_backend > > On Wed, Nov 27, 2013 at 6:56 PM, Stefan Priebe - Profihost AG > wrote: >> the performance comparisions are very impressive: >> >> https://github.com/facebook/rocksdb/wiki/Performance-Benchmarks >> >> Stefan >> >> Am 27.11.2013 11:55, schrieb Stefan Priebe - Profihost AG: >>> Hi, >>> >>> while googles leveldb was too slow for facebook they created rocksdb >>> (http://rocksdb.org/) may be interesting for Ceph? It's already >>> production quality. >>> >>> Greets, >>> Stefan >>> _______________________________________________ >>> ceph-users mailing list >>> ceph-users@lists.ceph.com >>> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com >>> >> -- >> 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 > > >