发自我的 iPhone 在 2013-7-23，4:35，"Charles 'Boyo" <charlesboyo@xxxxxxxxx<mailto:charlesboyo@xxxxxxxxx>> 写道： Hi, On Mon, Jul 22, 2013 at 2:08 AM, Chen, Xiaoxi <xiaoxi.chen@xxxxxxxxx<mailto:xiaoxi.chen@xxxxxxxxx>> wrote: Hi, > Can you share any information on the SSD you are using, is it PCIe connected? Depends, if you use HDD as your OSD data disk, a SATA/SAS SSD is enough for you. Instead of Intel 520, I would like to suggest you use the Intel DCS3700 since it provide better durability for write. Since a DCS3700 can provide 400~500MB/s for write and HDD can only have ~100MB/s ,it’s safe for a DCS3700 to provide journal for 4~5 HDDs. Indeed, I am using 2TB SATA 7200 drives in the OSDs, but unable to make room for the SSD, not even for a 2.5" drive. Considering using a mSATA to PCIe adapter with a SATA III mSATA SSD. Any thoughts on what to expect from this combination? And , if you have some insight/assumption on your workload, say “ I don’t care throughtput at all , all my workload doing random access”. With such assumption , you can have very high SSD:HDD ratio, 8:1 or even 10:1 will also be fine My workload will be read and write heavy, mostly random I/O (mail servers, OpenStack VMs, multi-node video recording and streaming), I'm afraid this will kill smaller SSDs too quickly to be sustainable, not so? I would say NO if you use enterprise level SSD, for example DCS 3700, it could provide PB level durability for 4K random write(the more random the workload is, the larger the write amplify tatio ,thus less durability) Charles _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
 |