Content addressable storage in legacy systems
A CAS data storage system replicates data on a non-CAS storage device. The CAS storage device recognizes duplicate data and stores the data only once, whereas the non-CAS device does not recognize duplication of data and requires full storage of the data. The CAS data storage device saves on redundant data transfer by transferring, in the case of duplicate data, the address of a primary location at which the data is stored and the address of the current duplication. The CAS data storage system includes a hash→address table for this purpose. The non-CAS storage device then copies its own data from the primary location into the current location.
1. A data storage system with backup protection using data replication, the system comprising:
at least one source CAS data storage space; and
a destination data storage space not having or not known to have CAS capabilities;
said source CAS storage space comprising:
an input for receiving data and corresponding addresses for storage in the source data storage space;
a hash key generator for generating a hash key of a data item;
a hash key to address storage table configured to store respective hash keys in the source storage space in association with corresponding logical addresses, to retrieve said addresses following input of a hash key;
an address to hash key storage table connecting each logical address to a respective corresponding hash key, to retrieve respective hash keys following input of a respective address;
a hash key inspection unit configured to compare stored hash keys against hash keys generated from further data items arriving at the source storage space to determine whether respective further data items are already stored, the source storage system configured to transfer, to said destination storage space, respective further data items together with corresponding addresses if no match is made to a stored hash key, and if a match is made to a stored hash key, then to transfer to said destination storage space both said address corresponding to respective further data items and respective addresses corresponding to said stored hash keys, without transferring said respective further data item; and
said destination data storage space being configured to use said address corresponding to said stored hash key to obtain said respective further data item and to copy said respective further data item to said address corresponding to said data item.
2. The data replication system of claim 1 , wherein said source data storage system is configured to obtain said address corresponding to said stored hash key from said hash to address table for said transferring.
3. The data replication system of claim 1 , wherein said address corresponding to said hash and said address corresponding to said further data are respectively logical addresses.
4. The data replication system of claim 1 , wherein said address corresponding to said hash and said address corresponding to said further data are respectively logical addresses within a logical address range and said hash key to address storage table comprises all of said logical address range.
5. The data replication system of claim 1 , wherein said address corresponding to said hash and said address corresponding to said further data are respectively logical addresses within a logical address range and said hash key to address storage table comprises only parts of said logical address range currently in use.
6. A source CAS data storage device for storing incoming data and maintaining a replica of said stored data at a destination data storage space not having or not known to have CAS capabilities, the source CAS data storage device comprising:
a source CAS data storage space;
an input for receiving data and corresponding addresses for storage in the source data storage space;
a hash key generator for generating a hash key of a data item;
a hash key to address storage table configured to store respective generated hash keys in the source storage space and to link respective hash keys to retrieve corresponding addresses;
an address to hash key storage table connecting each hash key to an address at which corresponding data is placed to retrieve respective hash keys on entry of corresponding addresses; and
a hash key inspection unit configured to compare prestored hash keys against hash keys generated from further data items arriving at the source storage space to determine whether respective further data items are already stored, the source storage system configured to transfer, to said destination storage space, respective further data items together with corresponding addresses if no match is made to a prestored hash key, and if a match is made to a prestored hash key, then to transfer to said destination storage space both said address corresponding to said further data item and said address corresponding to said hash key, without transferring said respective further data item.
7. A data replication method comprising:
receiving data and corresponding addresses for storage in a source CAS data storage space;
generating a hash key of respective received data items;
generating a hash key to address storage table to store generated hash keys in the source CAS storage space to link respective hash keys to corresponding addresses;
comparing prestored hash keys against hash keys generated from further data items arriving at the source storage space to determine whether respective further data items are already stored; and
replicating, at a destination storage space, data stored at said source storage space by transferring, to said destination storage space, respective further data items together with corresponding addresses if no match is made to a prestored hash key, and if a match is made to a prestored hash key, then transferring to said destination storage space both said address corresponding to said data item and said address corresponding to said hash key, without transferring said respective further data item.
8. The method of claim 7 , further comprising, at said destination storage space, using said address corresponding to said hash key to obtain said respective further data item and to copy said respective further data item to said address corresponding to said data item.
9. The method of claim 7 , comprising generating at said source data storage system, an address to hash table together with said hash to address table.
10. The method of claim 7 , wherein said address corresponding to said hash and said address corresponding to said further data are respectively logical addresses.
11. The method of claim 7 , wherein said address corresponding to said hash and said address corresponding to said further data are respectively logical addresses within a logical address range and said hash key to address storage table comprises all of said logical address range.
12. The method of claim 7 , wherein said address corresponding to said hash and said address corresponding to said further data are respectively logical addresses within a logical address range and said hash key to address storage table comprises only parts of said logical address range currently in use.