A Beijing real estate company, an HP Dl380g7 server, because of physical damage to multiple disks, resulting in raid crashes, data loss.
The HP Dl380g7 server runs a Kingdee EAS system with a background of more than 40 GB of Oracle database, and the database data is very important.
The HP Dl380g7 server is plugged into 8 146GB SAS disks, and each 4 disk has a set of RAID5, and a total of two sets of RAID5 are made into a RAID0, a total of 8 disk RAID50.
8 disk has 3 disk damage, and offline in the raid, beyond the raid redundancy range, is the main reason for data loss.
The customer was introduced to the Beijing security Number Cloud and Technology Co., Ltd. Beijing security Number Cloud and Technology Co., Ltd. 8 years of enterprise-level data recovery experience, for a variety of servers, a variety of storage and a variety of databases have a very unique technology, and self-developed VMWare vsphere-esxi VMFS data Recovery Software (Vmxdb for VMware VMFS Recovery), AIX data Recovery Software (VMXDB for Aix Recovery), HPUX data Recovery Software (VMXDB for Hpunix Recovery) and a variety of data recovery software.
Beijing Wouldl Cloud and the beginning of the detection of all the disk, found that there are two of disks have a large number of bad, and a disk of the head damage, need to open processing. After processing the physical damage to the disk, the data of all disks is mirrored.
Then through the underlying data of each disk to analyze the order of each disk in the original RAID and stripe size information, and then use the virtual RAID technology, the software to simulate the original RAID card on the disk organization structure, virtual reorganization raid, and then all the data recovered.
Customer's original disk:
Virtual reorganization raid, you can see the data:
Then, because the disk has a large number of non-repairable bad lanes, resulting in the Oracle database has a large number of bad blocks, there is an important data file with no head, the head has 10 blocks, 80KB, all 0 bytes.
Try to mount the database in the Beijing security cloud and the Oracle environment within the company:
Only repair the Oracle database, refactor a usable file header, and fix the bad block.
After repairing the head data, the detected bad block:
After the repair, mount the database again, OK.
Pull up the repaired database with the customer's machine and verify with the customer's application, no problem.
Http://blog.ifeng.com/article/44890758.html
Http://blog.ifeng.com/article/44890830.html
Http://blog.ifeng.com/article/44890961.html
Http://site.leshou.com/s/27013070.html
Http://site.leshou.com/s/27013553.html
Http://site.leshou.com/s/27013729.html
Http://site.leshou.com/s/27014199.html
Http://site.leshou.com/s/27014515.html
Http://blog.ifeng.com/article/44891667.html
Http://www.360yue.cn/news/show-82834.html
Http://www.360yue.cn/news/show-82836.html
Http://www.360yue.cn/news/show-82837.html
Http://shiyou.huangye88.com/xinxi/81336371.html
Http://shiyou.huangye88.com/xinxi/81336836.html
Http://shiyou.huangye88.com/xinxi/81337163.html
Http://mt.sohu.com/20160530/n452037737.shtml
Http://mt.sohu.com/20160530/n452039008.shtml
Http://mt.sohu.com/20160530/n452039295.shtml
Http://www.99inf.com/jinrong/qhqq/757503.html
Http://www.99inf.com/jinrong/gpzq/757512.html
Http://mt.sohu.com/20160530/n452041589.shtml
Http://www.99inf.com/jinrong/qhqq/757527.html
Http://www.docin.com/p-1608795503.html
Http://www.docin.com/p-1608798502.html
Http://www.docin.com/p-1608804191.html
Http://blog.dqccc.com/s/article_6561658.html
Http://blog.dqccc.com/s/article_6561723.html
Http://blog.sina.com.cn/s/blog_7362e71e0102wh34.html
Http://blog.ifeng.com/article/44899854.html
Http://blog.sina.com.cn/s/blog_7362e71e0102wh35.html
Http://blog.ifeng.com/article/44899945.html
Http://blog.ifeng.com/article/44900031.html
Http://blog.sina.com.cn/s/blog_7362e71e0102wh37.html
Http://blog.ifeng.com/article/44900154.html
hp-dl380 data Recovery and Kingdee EAS data recovery Oracle Repair process