A record is a particular structure that composes a reference to your information that makes it less demanding to gaze upward. In Postgres it is a duplicate of the thing you wish to file joined with a reference to the genuine information area. While getting to information, Postgres will either utilize some type of a file on the off chance that it exists or a successive output. A successive sweep is the point at which it looks over the greater part of the information before giving back the outcomes. HDD (Hard Disk Drive) has a wide crevice between the execution of consecutive I/O and that of irregular I/O inferable from its mechanical parts. Because of this reality, DBMS more often than not lean toward a full table sweep to a record filter aside from when selectivity is sufficient low to exploit list check. Here, an advancement of the list sweep is actualized, called streak mindful list check by joining two ideas in PostgreSQL. An idea is a sorted list filter that sweeps tuples all together of record ids.
[1]
Karsten Schwan,et al.
Unified address translation for memory-mapped SSDs with FlashMap
,
2015,
2015 ACM/IEEE 42nd Annual International Symposium on Computer Architecture (ISCA).
[2]
Yun Wang,et al.
An efficient hybrid join algorithm: a DB2 prototype
,
1991,
[1991] Proceedings. Seventh International Conference on Data Engineering.
[3]
Xiaodong Zhang,et al.
Essential roles of exploiting internal parallelism of flash memory based solid state drives in high-speed data processing
,
2011,
2011 IEEE 17th International Symposium on High Performance Computer Architecture.
[4]
Min Wang,et al.
Efficient Parallel Processing of Distance Join Queries Over Distributed Graphs
,
2015,
IEEE Transactions on Knowledge and Data Engineering.