From owner-freebsd-fs@FreeBSD.ORG Thu Apr 9 03:18:52 2009 Return-Path: Delivered-To: freebsd-fs@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 59F16106566C for ; Thu, 9 Apr 2009 03:18:52 +0000 (UTC) (envelope-from hali@datapipe.com) Received: from EXFESMQ01.datapipe-corp.net (exfesmq01.datapipe-corp.net [64.106.130.69]) by mx1.freebsd.org (Postfix) with ESMTP id 259DC8FC16 for ; Thu, 9 Apr 2009 03:18:52 +0000 (UTC) (envelope-from hali@datapipe.com) Received: from datapipe.net (192.168.128.20) by EXFESMQ01.datapipe-corp.net (64.106.130.71) with Microsoft SMTP Server (TLS) id 8.1.340.0; Wed, 8 Apr 2009 23:17:33 -0400 Date: Wed, 8 Apr 2009 23:18:51 -0400 From: Hussain Ali To: Message-ID: <20090409031851.GE6052@datapipe.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Disposition: inline User-Agent: mutt-ng/devel-r804 (Linux) Content-Transfer-Encoding: quoted-printable Subject: Re: ZFSKnownProblems - needs revision? X-BeenThere: freebsd-fs@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Filesystems List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 09 Apr 2009 03:18:52 -0000 > Ivan Voras wrote: > > >>* Are the issues on the list still there? >>* Are there any new issues? >>* Is somebody running ZFS in production (non-trivial loads) with >>success? What architecture / RAM / load / applications used? >>* How is your memory load? (does it leave enough memory for other >>services) I have a storage server its constantly heavy writing and reading (at times) though not with high concurrency: # df -h Filesystem Size Used Avail Capacity Mounted on /dev/ufs/rootfs 19G 395M 17G 2% / devfs 1.0K 1.0K 0B 100% /dev /dev/ufs/tmp 4.8G 20K 4.5G 0% /tmp /dev/ufs/usr 19G 2.9G 15G 16% /usr /dev/ufs/var 15G 7.5G 5.8G 56% /var backupstorage 94T 80T 13T 86% /backupstorage # cat /etc/sysctl.conf # $Id: sysctl.conf,v 1.3 2009/04/09 03:06:31 hali Exp root $ security.bsd.see_other_uids=3D0 net.inet.tcp.blackhole=3D2 net.inet.udp.blackhole=3D1 net.inet.icmp.icmplim=3D50 net.inet.tcp.sendspace=3D524288 net.inet.tcp.recvspace=3D524288 net.inet.ip.intr_queue_maxlen=3D2048 net.inet.ip.intr_queue_drops=3D4096 kern.ipc.maxsockbuf=3D2097152 kern.ipc.somaxconn=3D8096 kern.maxfiles=3D443808 vfs.hirunningspace=3D4194304 vfs.ufs.dirhash_maxmem=3D4194304 vfs.lookup_shared=3D1 # cat /boot/loader.conf # $Id: loader.conf,v 1.4 2009/04/09 03:07:40 hali Exp root $ isp_load=3D"YES" ispfw_load=3D"YES" isp_2400_load=3D"NO" vm.kmem_size_max=3D"1073741824" vm.kmem_size=3D"1073741824" vfs.zfs.prefetch_disable=3D1 vfs.zfs.arc_max=3D"786M" kern.maxvnodes=3D"50000 # zpool iostat 3 capacity operations bandwidth pool used avail read write read write ------------- ----- ----- ----- ----- ----- ----- backupstorage 80.4T 14.9T 2 355 186K 38.7M backupstorage 80.4T 14.9T 0 316 0 31.7M backupstorage 80.4T 14.9T 0 99 0 12.2M backupstorage 80.4T 14.9T 0 164 0 15.7M backupstorage 80.4T 14.9T 0 225 0 22.0M I have another another in another dc, but less capacity: backupstorage 56T 32T 24T 57% /backupstorage Both are the following: HP ProLiant DL385 G2 8GB RAM dual dual core AMD 2.2Ghz cpus 3 x Nexsan SataBEAST for the san. Inbound about 900Mb/s, uptime has generally been 3-4 months before increasing the ZFS arc/KVM sizes. I should just max it out but its relatively stable. Am looking for ZFSv8+ for L2Arc and separate ZIL. Load averages about ~ 1.0 . My wish list would be KVM support fot 64GB ARC, fusion-IO driver support for the ZIL, version 8 of ZFS in FreeBSD 7.2, active multipath, etc, etc.. It works, its stable, its production, but its not like i am cvsuping ports tree 100 times concurrently. -- -hussain This message may contain confidential or privileged information. If you ar= e not the intended recipient, please advise us immediately and delete this = message. See http://www.datapipe.com/emaildisclaimer.aspx for further info= rmation on confidentiality and the risks of non-secure electronic communica= tion. If you cannot access these links, please notify us by reply message a= nd we will send the contents to you.