Methods and systems for recovering meta-data in a cache memory after a corruption event
A method for recovering meta-data that has been subjected to corruption is disclosed for a cache memory. Each table entry in the cache memory has an in-cache indicator for indicating whether the data unit associated therewith is stored in a cache-slot in the cache memory. A first review is conducted of the table entries for identifying in-cache table entries (those that are associated with cache-slots in the cache memory), performing a correction attempt to ensure that the meta-data in an in-cache table entry are validated and corrected, if necessary, and ensuring that the in-cache indicator is set for an in-cache table entry. A second review is conducted of the table entries to ensure that the in-cache indicator is not set for any table entry that is not an in-cache table entry or for any in-cache table entry for which the correction attempt was unsuccessful.
1. For a cache memory having meta-data organized into a plurality of table entries, and having data units stored in a plurality of cache-slots, with each of the cache-slots being arranged to be associated with one of the table entries; a method for recovering meta-data that have been subjected to corruption, comprising
providing each of the table entries with an in-cache indicator to indicate that the data unit associated therewith is stored in the cache memory;
conducting a first review of the table entries, further comprising
identifying in-cache table entries comprising the table entries that are associated with cache-slots in the cache memory,
performing a correction attempt on each of the in-cache table entries to ensure that the meta-data in each of the in-cache table entries are validated and corrected, if necessary, and
ensuring that the in-cache indicator is set for each of the in-cache table entries; and conducting a second review of the table entries, further comprising
identifying not-in-cache table entries comprising the table entries that are not associated with cache-slots in the cache memory,
identifying invalid table entries comprising the in-cache table entries for which the correction attempt was unsuccessful;
ensuring that the in-cache indicator is not set for any of the not-in-cache table entries and for any of the invalid table entries.
2. The method of claim 1 , wherein conducting the first review further comprises conducting a seriatim review of each of the cache-slots, the seriatim review further comprising
selecting a first cache-slot;
identifying a first table entry that is associated with the first cache-slot;
determining whether the first table entry is in need of repairing; and
performing a correction attempt on the first table entry to correct and verify validity of the meta-data in the first table entry when the first table entry is determined to be in need of repairing.
3. The method of claim 2 ,
wherein each of the table entries has a stored meta-data protection value and each of the cache-slots has a backup table entry comprising a backup copy of an associated table entry comprising a table entry associated therewith; and
further comprising conducting a data protection correction process to verify validity of a data protection portion of the first table entry by
calculating a test meta-data protection value from the first table entry;
comparing the test meta-data protection value and the stored meta-data protection value; and
when the test meta-data protection value and the stored meta-data protection value do not match, determining whether the backup table entry is in need of repairing; and, if the backup table is not in need of repairing, replacing the first table entry with the backup table entry.
4. The method of claim 2 ,
wherein the first table entry has a cache-slot association link comprising a pointer to one of the cache-slots to link the first table entry to the one of the cache-slots; and
further comprising conducting a lookup table correction process to verify validity of a lookup table portion of the first table entry; wherein the lookup table correction process further comprises confirming the cache-slot association link from the first table entry to the first cache-slot by
using the cache-slot association link to access the one of the cache-slots; and
replacing the cache-slot association link with a pointer to the first cache-slot when the cache-slot association link does not access the first cache-slot.
5. The method of claim 1 , wherein conducting the second review further comprises conducting a seriatim review of each of the table entries, the seriatim review further comprising selecting a first table entry;
determining whether the first table entry is in need of repairing;
ensuring that the in-cache indicator is unset for the first table entry when the first table entry is in need of repairing.
6. The method of claim 5 , wherein each of the table entries has a stored meta-data protection value; and
further comprising conducting a data protection validation process to verify validity of a data protection portion of the first table entry by
calculating a test meta-data protection value from the first table entry;
comparing the test meta-data protection value and the stored meta-data protection value; and
ensuring that the in-cache indicator is unset for the first table entry when the test meta-data protection value and the stored meta-data protection do not match.
7. The method of claim 5 ,
wherein the first table entry has a cache-slot association link comprising a pointer to one of the cache-slots to link the first table entry to the one of the cache-slots;
further comprising conducting a lookup table validation process to verify validity of a lookup table portion of the first table entry; wherein the lookup table validation process further comprises confirming the cache-slot association link from the first table entry to the first cache-slot by
using the cache-slot association link to access the one of the cache-slots; and
ensuring that the in-cache indicator is unset for the first table entry when the cache-slot association link does not access the first cache-slot.
8. The method of claim 1 , wherein each of the table entries has a data protection process associated therewith, and further comprising:
suspending the data protection process for any of the table entries for which the in-cache indicator is not set; and
deferring re-establishing the data protection process for a first table entry for which the in-cache indicator is not set until issuance of a request for an I/O operation therefor.
9. The method of claim 1 , wherein each of the table entries has a data protection process associated therewith, and further comprising:
suspending the data protection process for any of the not-in-cache table entries; and
deferring re-establishing the data protection process for a first not-in-cache table entry until issuance of a request for an I/O operation therefor.
10. The method of claim 1 , wherein each of the table entries has a data protection process associated therewith, and further comprising:
suspending the data protection process for any of the invalid table entries; and
deferring re-establishing the data protection process for a first invalid table entry until issuance of a request for an I/O operation therefor.
11. A computer program product residing on a computer readable storage medium for recovering meta-data for a computer system that have been subjected to corruption,
the computer system having a cache memory with the meta-data organized into a plurality of table entries, and with data units stored in a plurality of cache-slots, each of the cache-slots being arranged to be associated with one of the table entries;
the computer program product comprising instructions for causing a computer to:
conduct a first review of the table entries, further comprising
identifying in-cache table entries comprising the table entries that are associated with cache-slots in the cache memory,
performing a correction attempt on each of the in-cache table entries to ensure that the meta-data in each of the in-cache table entries are validated and corrected, if necessary, and
ensuring that the in-cache indicator is set for each of the in-cache table entries; and
conduct a second review of the table entries, further comprising
identifying not-in-cache table entries comprising the table entries that are not associated with cache-slots in the cache memory,
identifying invalid table entries comprising the in-cache table entries for which the correction attempt was unsuccessful;
ensuring that the in-cache indicator is not set for any of the not-in-cache table entries and for any of the invalid table entries.
12. The computer program product of claim 11 , wherein the instructions for causing a computer to conduct a first review of the table entries further comprises instructions to cause a computer to conduct a seriatim review of each of the cache-slots, the seriatim review further comprising
selecting a first cache-slot;
identifying a first table entry that is associated with the first cache-slot;
determining whether the first table entry is in need of repairing; and
performing a correction attempt on the first table entry to correct and verify validity of the meta-data in the first table entry when the first table entry is determined to be in need of repairing.
13. The computer program product of claim 11 , wherein the instructions for causing a computer to conduct a second review of the table entries further comprises instructions to cause a computer to conduct a seriatim review of each of the table entries, the seriatim review further comprising
selecting a first table entry;
determining whether the first table entry is in need of repairing;
ensuring that the in-cache indicator is unset for the first table entry when the first table entry is in need of repairing.
14. The computer program product of claim 11 , wherein each of the table entries has a data protection process associated therewith, and further comprising instructions for causing a computer to:
suspend the data protection process for any of the table entries for which the in-cache indicator is not set; and
defer re-establishing the data protection process for a first table entry for which the in-cache indicator is not set until issuance of a request for an I/O operation therefor.
15. A data storage system comprising:
a cache memory having meta-data organized into a plurality of table entries, and having data units stored in a plurality of cache-slots, each of the cache-slots being arranged to be associated with one of the table entries; and
a processor coupled to the cache memory and operable to recover meta-data that have been subjected to corruption, the processor further operable to conduct a first review of the table entries, further comprising
identifying in-cache table entries comprising the table entries that are associated with cache-slots in the cache memory,
performing a correction attempt on each of the in-cache table entries to ensure that the meta-data in each of the in-cache table entries are validated and corrected, if necessary, and
ensuring that the in-cache indicator is set for each of the in-cache table entries; and
conduct a second review of the table entries, further comprising
identifying not-in-cache table entries comprising the table entries that are not associated with cache-slots in the cache memory,
identifying invalid table entries comprising the in-cache table entries for which the correction attempt was unsuccessful;
ensuring that the in-cache indicator is not set for any of the not-in-cache table entries and for any of the invalid table entries.
16. The data storage system of claim 15 , wherein the processor being operable to conduct a first review of the table entries further comprises the processor being operable to conduct a seriatim review of each of the cache-slots, the seriatim review further comprising
selecting a first cache-slot;
identifying a first table entry that is associated with the first cache-slot;
determining whether the first table entry is in need of repairing; and
performing a correction attempt on the first table entry to correct and verify validity of the meta-data in the first table entry when the first table entry is determined to be in need of repairing.
17. The data storage system of claim 15 , wherein the processor being operable to conduct a second review of the table entries further comprises the processor being operable to conduct a seriatim review of each of the table entries, the seriatim review further comprising selecting a first table entry;
determining whether the first table entry is in need of repairing;
ensuring that the in-cache indicator is unset for the first table entry when the first table entry is in need of repairing.
18. The data storage system of claim 15 , wherein each of the table entries has a data protection process associated therewith, and further comprising the processor being operable to:
suspend the data protection process for any of the table entries for which the in-cache indicator is not set; and
defer re-establishing the data protection process for a first table entry for which the in-cache indicator is not set until issuance of a request for an I/O operation therefor.