Computing device replication using file system change detection methods and systems

Abstract

Computing device file system change detection and change replication methods and systems are described herein. In one embodiment, the change detection is accomplished by comparing two master file tables of two different versions of the same file system and inferring file system name space change, file system metadata change and file data change. The method includes creating a file system name space change log and file data change log for any objects associated with the changes to the file system. The method also includes replicating these change logs to the replication server and applying the change logs to the data store on replication server.

What is claimed is: 
     
         1 . A method, comprising:
 a replication agent obtaining a first master file table for a client computing device;   obtaining a second master file table for the client computing device;   comparing the first master file table with the second master file table to determine changes to the first master file table relative to the second master file table that are indicative of changes to a file system of the client computing device; and   creating a file system name space change log and file data change log for the file system, comprising of directories and files.   
     
     
         2 . The method according to  claim 1 , further comprising replicating the changed inodes on the replication appliance to a cloud data store. 
     
     
         3 . The method according to  claim 1 , wherein detecting changed inodes comprises detecting a change in an inode generation number of an inode of the first set of inodes that is indicative of reuse of the inode for a new object. 
     
     
         4 . The method according to  claim 3 , wherein if reuse is detected, the method further comprises creating an entry in a log that records creation of the new object and deletion of an old object which was replaced by the new object. 
     
     
         5 . The method according to  claim 1 , wherein the changes are determined from examining log sequence numbers within the first and second master file tables. 
     
     
         6 . The method according to  claim 1 , wherein detecting changed inodes comprises detecting a modified time entry of an inode of the first set of inodes which is indicative of a change in an object associated with the inode. 
     
     
         7 . The method according to  claim 1 , wherein detecting changed inodes comprises detecting a new name for an inode of the first set of inodes, and further logging the new name in a log. 
     
     
         8 . The method according to  claim 1 , wherein detecting changed inodes comprises detecting a change to a parent identifier for an inode of the first set of inodes which is indicative of movement of the inode within the file system. 
     
     
         9 . The method according to  claim 1 , further comprising copying metadata and namespace changes for the file system by copying only the disk blocks that store the metadata and namespace information (cloned shell volume) to the replication appliance. 
     
     
         10 . The method according to  claim 1 , wherein translating the changed inodes into namespace locations comprises obtaining file identifiers of the changed inodes and performing file path identification to obtain a file path for each of the changed inodes based on the file identifiers. 
     
     
         11 . The method according to  claim 1 , further comprising converting the file data change log to a refined log by:
 performing a tree walk of a directory in a file system for the client computing device for a directory rename operation of a directory or file move operations of a file;   adding a create record for child files of the directory or the object;   delete any modified file record if the file; and   wherein the refined log comprises file change records that are an incremental change or a full backup of the client computing device.   
     
     
         12 . A system, comprising:
 a cloud data store; and   a replication appliance associated with a client computing device, the replication appliance comprising a processor and memory, wherein the processor executes logic stored in memory to:   obtain a first master file table of a file system for a client computing device;   obtain a second master file table of the file system for the client computing device;   compare the first master file table with the second master file table to determine changes to the first master file table relative to the second master file table that are indicative of changes to a file system of the client computing device;   create a file data change log for any objects associated with the changes to the file system; and   replicate blocks associated with the objects on the replication appliance to the cloud data store.   
     
     
         13 . The system according to  claim 12 , wherein the processor further executes the logic to replicate metadata of the file system onto the replication appliance. 
     
     
         14 . The system according to  claim 13 , wherein the processor further executes the logic to convert the file data change log to a refined log by:
 performing a tree walk of a directory in a file system for the client computing device for a directory rename operation of a directory or file move operations of a file;   adding a record for child files of the directory or the object;   delete any modified file record if the file; and   wherein the refined log comprises file change records that are an incremental change or a full backup of the client computing device.   
     
     
         15 . The system according to  claim 12 , wherein the changes include file name changes, parent identifier changes, sequence number changes, time signature changes, log sequence changes, and any combinations thereof. 
     
     
         16 . A method, comprising:
 obtaining pairs of master file table (MFT) entries in file identifier numerical order, the pairs MFT entries being obtained from a first snapshot of a client computing device and a second snapshot of the client computing device;   creating a record for a second MFT entry of a pair of MFT entries if a first of the pair of MFT entries is invalid and the second MFT entry is valid;   deleting a record for a first MFT entry of the pair of MFT entries if the first of the pair of MFT entries is valid and the second MFT entry is invalid;   wherein if both the first MFT entry and the second MFT entry are valid and generation numbers of the first and second MFT entries refer to a same object, the method further comprises obtaining name attributes for each of the first and second MFT entries;   further wherein:   if name attributes from the first MFT entry that do not exist in the second MFT entry, an unlink record is generated; and if name attributes from the second MFT entry that do not exist in the first MFT entry, a link record is generated.