Case:
HP server, UNIX system, important Oracle database was lost by the engineer, losing all data tables, UNDO, LOG, etc.The volume file system is VXFS, the volume size is about 500g, and the total data is about 230g.Disaster is extremely important.The user is very cautious and at all cost, but most companies cannot provide solutions.Even most data recovery companies do not have data recovery cases and experience under HP-UX.After the user fully describes the phenomenon, we provide at least two solutions.Because there are many similar cases before, and the user does not hesitate, there is no need to conceal the failure, so the conclusion we give is that the data will be successfully recovered at 100%under the premise of spending more time and energy.
Solution:
After the engineer logged in to the HP-UX system remotely, the DD operation of the fault volume was output to another volume.It is transmitted to the Windows platform through FTP and implemented according to the scheme. The independent data recovery software successfully analyzes other file node tables outside the UNDO1, and extracts all the files according to the node.According to the implementation of the second plan, all the files are successfully restored. Except for UNDO1, comparison with the results of scheme 1, 1 byte is not bad. Therefore, it is determined that the remaining files are 100%correct except UNDO1.Because UNDO1 does not affect the data, and it is correct according to the conclusion.Therefore, data resumption is successful.After the customer restores all the database back to the HP-UX system, the service is successfully started, and the data is intact.
