HDS产品介绍与技术解决方案.pptVIP

* * Hitachi allows for volumes to be powered off when not be accessed making a very cost effective platform for backup or archive data used infrequently. Saves on electricity as the disks do not use power when powered off and also reduces cooling costs (fewer BTUs when off). * * * * * * * * * Initial view – the physical layout of storage Note: Volume color corresponds to storage system where the disks reside Striped volumes represent unused/available volumes Cache is shown in the USP (partially hidden by internal, orange, volumes in this slide, but will be visible in subsequent slides) * * Allows data to be moved from SATA to FC (or FC to SATA) while the storage is active (no server volume remount required). Allows LUNs to be moved from RAID5 to RAID6 or different sized RAID groups (i.e. 4D + 1P to 7D + 1P). This feature will be available in Q3 2006 * 对资料互通性在虚拟化的过程里面可能会随着互阶存储，在低阶继续访问情况下，往重要的存储设备搬，也可以把次级存储往第三级搬，也可以往外搬做备份的工作，因此在虚拟化存储池里面就可以游刃有余的资料搬迁，而且不需要应用系统过点时间操作，因为这都是24小时的运作。 [ 16:14 ] * Now, we can also [logically] see that we can pool available storage. (A basic USP benefit.) * * 【徐志钧】另外随着应用系统增加，随着各个应用系统需要，要复制来复制去，作为资料的复制方案，我们可以把异阶存储，把性能最高的资料也可以通过系统快造的功能快速录制底下二级的两台存储设备上，也可以从二阶搬到三阶也可以在线在同一个机器里面通过虚拟化的引擎、平台做资料的搬迁，这是何等的神奇，可以把这资料做很有效的迁移。 [ 16:15 ] * * * 介绍要点 数据的复制过程，1，2，3，4，５ I/O响应时间惩罚，远程I/O未完成之前，不会回应主机I/O完成 优点： 数据完全同步，可用 缺点： 距离限制，理论极限值100km；距离越长，时延越大 双倍的I/O响应时间，实时的Random I/O可以忍受，批量Sequential I/O无法忍受 Detail explanation： Note that commitment is made when the storage cache receive/secure the data. In synchronous mode, data are store in the cache first. The primary storage will not de-stage the data (write to disk) until it receive the acknowledgement from the secondary storage cache. All vendor can do this. This has a distance limitation of 40KM to 100KM. It will have an impact on response time, for data must be “committed” by the remote site storage system before getting back to the end user originated the transaction. Going for synchronous mode will also prolong