From mboxrd@z Thu Jan 1 00:00:00 1970 From: Alexandre DERUMIER Subject: Re: Memstore performance improvements v0.90 vs v0.87 Date: Sat, 21 Feb 2015 10:02:53 +0100 (CET) Message-ID: <1725252187.1574758.1424509373883.JavaMail.zimbra@oxygem.tv> References: <3649A15A2562B54294DE14BCE5AC79120AB30A5D@FMSMSX106.amr.corp.intel.com> <54C7FD1C.40406@redhat.com> <54E6F96B.9080202@ubuntu.com> <54E754DA.6060703@redhat.com> <252858605.1545602.1424448205171.JavaMail.zimbra@oxygem.tv> <54E75CFE.70601@redhat.com> <893632926.1573877.1424506229039.JavaMail.zimbra@oxygem.tv> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Received: from mailpro.odiso.net ([89.248.209.98]:52910 "EHLO mailpro.odiso.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751754AbbBUJC4 convert rfc822-to-8bit (ORCPT ); Sat, 21 Feb 2015 04:02:56 -0500 In-Reply-To: <1429598219.1574757.1424509359439.JavaMail.zimbra@oxygem.tv> Sender: ceph-devel-owner@vger.kernel.org List-ID: To: Mark Nelson Cc: Blair Bethwaite , James Page , ceph-devel , Stephen L Blinick , Jay Vosburgh , Colin Ian King , Patricia Gaughen , Leann Ogasawara >>One Other thing you could also check, is if transparent hugepages is = enabled or not. >>It's known to slowdown some applications, like databases for example.= (So maybe memstore too) >> >>#cat /sys/kernel/mm/transparent_hugepage/enabled >> always madvise [never] This is disabled by default is you use the network-latency tuned profil= e https://developerblog.redhat.com/2015/02/11/low-latency-performance-tun= ing-rhel-7/ ----- Mail original ----- De: "aderumier" =C3=80: "Mark Nelson" Cc: "Blair Bethwaite" , "James Page" , "ceph-devel" , "Stephen L= Blinick" , "Jay Vosburgh" , "Colin Ian King" , "Patricia G= aughen" , "Leann Ogasawara" Envoy=C3=A9: Samedi 21 F=C3=A9vrier 2015 09:10:29 Objet: Re: Memstore performance improvements v0.90 vs v0.87 >>At least at first glance it doesn't look like it's enabled on either = our=20 >>RHEL7 or Ubuntu nodes. No numad running either.=20 Do you have disable autonuma balancing manually ?=20 because look at kernel config of both rhel7 and ubuntu, it's really see= m to be enabled by default:=20 CONFIG_ARCH_SUPPORTS_NUMA_BALANCING=3Dy=20 CONFIG_NUMA_BALANCING_DEFAULT_ENABLED=3Dy=20 CONFIG_NUMA_BALANCING=3Dy=20 >>No numad running either.=20 AFAIK, numad was used by RHEL6, and auto-numabalancing is used since RH= EL7.=20 One Other thing you could also check, is if transparent hugepages is en= abled or not.=20 It's known to slowdown some applications, like databases for example. (= So maybe memstore too)=20 #cat /sys/kernel/mm/transparent_hugepage/enabled=20 always madvise [never]=20 ----- Mail original -----=20 De: "Mark Nelson" =20 =C3=80: "aderumier" =20 Cc: "Blair Bethwaite" , "James Page" , "ceph-devel" , "Stephen L= Blinick" , "Jay Vosburgh" , "Colin Ian King" , "Patricia G= aughen" , "Leann Ogasawara" =20 Envoy=C3=A9: Vendredi 20 F=C3=A9vrier 2015 17:12:46=20 Objet: Re: Memstore performance improvements v0.90 vs v0.87=20 On 02/20/2015 10:03 AM, Alexandre DERUMIER wrote:=20 >>> http://rhelblog.redhat.com/2015/01/12/mysteries-of-numa-memory-mana= gement-revealed/=20 >>> It's possible that this could be having an effect on the results.=20 >=20 > Isn't auto numa balancing enabled by default since kernel 3.8 ?=20 No idea, I'm behind the times on what's been going on with numa tuning.= =20 >=20 > it can be checked with=20 >=20 > cat /proc/sys/kernel/numa_balancing=20 At least at first glance it doesn't look like it's enabled on either ou= r=20 RHEL7 or Ubuntu nodes. No numad running either.=20 --=20 To unsubscribe from this list: send the line "unsubscribe ceph-devel" i= n=20 the body of a message to majordomo@vger.kernel.org=20 More majordomo info at http://vger.kernel.org/majordomo-info.html=20 -- To unsubscribe from this list: send the line "unsubscribe ceph-devel" i= n the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html