Dynamic bloom filter adjustment for storage systems implementing a log structured merge tree architecture

A computer program product comprising a non-transitory processor-readable storage medium having stored therein program code of one or more software programs, wherein the program code when executed by at least one processing device causes the at least one processing device:

• to monitor input-output operations for a storage system, the storage system comprising a memory maintaining (i) a log structured merge tree cache for a log structured merge tree and (ii) a first set of one or more bloom filters, the log structured merge tree comprising a plurality of segments, each of the first set of one or more bloom filters being associated with a subset of the plurality of segments of the log structured merge tree, each bloom filter of the first set of one or more bloom filters having one or more parameters selected based at least in part on a workload pattern of the storage system determined in conjunction with addition of a given subset of the plurality of segments, associated with at least one bloom filter of the first set of one or more bloom filters, to the log structured merge tree;
• to determine, based at least in part on the monitored input-output operations, whether at least one of one or more bloom filter size adaptation trigger conditions have been met, the one or more bloom filter size adaptation trigger conditions comprising determining a current workload pattern of the storage system, the current workload pattern being determined based at least in part on monitoring a set of workload pattern characteristics of the monitored input-output operations;
• and
• responsive to determining that at least one of the one or more bloom filter size adaptation trigger conditions has been met, to adjust one or more parameters of a second set of one or more bloom filters maintained in the memory, the second set of one or more bloom filters being associated with one or more additional subsets of segments added to the log structured merge tree;
• wherein the second set of one or more bloom filters is different than the first set of one or more bloom filters;
• and
• wherein the one or more parameters of the first and second sets of one or more bloom filters comprise at least one bloom filter false positive ratio.

A method comprising:

• monitoring input-output operations for a storage system, the storage system comprising a memory maintaining (i) a log structured merge tree cache for a log structured merge tree and (ii) a first set of one or more bloom filters, the log structured merge tree comprising a plurality of segments, each of the first set of one or more bloom filters being associated with a subset of the plurality of segments of the log structured merge tree, each bloom filter of the first set of one or more bloom filters having one or more parameters selected based at least in part on a workload pattern of the storage system determined in conjunction with addition of a given subset of the plurality of segments, associated with at least one bloom filter of the first set of one or more bloom filters, to the log structured merge tree;
• determining, based at least in part on the monitored input-output operations, whether at least one of one or more bloom filter size adaptation trigger conditions have been met, the one or more bloom filter size adaptation trigger conditions comprising determining a current workload pattern of the storage system, the current workload pattern being determined based at least in part on monitoring a set of workload pattern characteristics of the monitored input-output operations;
• and
• responsive to determining that at least one of the one or more bloom filter size adaptation trigger conditions has been met, adjusting one or more parameters of a second set of one or more bloom filters maintained in the memory, the second set of one or more bloom filters being associated with one or more additional subsets of segments added to the log structured merge tree;
• wherein the second set of one or more bloom filters is different than the first set of one or more bloom filters;
• wherein the one or more parameters of the first and second sets of one or more bloom filters comprise at least one bloom filter false positive ratio;
• and
• wherein the method is performed by at least one processing device comprising a processor coupled to a memory.