TY - JOUR
T1 - On consistent reading of entire databases
AU - Lam, Kwok-Wa
AU - Lee, Victor C. S.
PY - 2006/4
Y1 - 2006/4
N2 - Many applications need to read an entire database in a consistent way. This global-reading of an entire database formulated as a global-read transaction (GRT) is not a trivial issue since it will cause a high degree of interference to other concurrent transactions. Conventional concurrency control protocols are obviously inadequate in handling the long-lived GRT. Previous studies [1], [3], [4] proposed additional tests, namely, the Color Test and the Shade Test, to handle conflicts between the GRT and update transactions. However, we discovered that both algorithms can bring about nonserializable schedules of transactions. We propose an enhanced algorithm directly built on the two algorithms to guarantee the serializability of transactions. © 2006 IEEE.
AB - Many applications need to read an entire database in a consistent way. This global-reading of an entire database formulated as a global-read transaction (GRT) is not a trivial issue since it will cause a high degree of interference to other concurrent transactions. Conventional concurrency control protocols are obviously inadequate in handling the long-lived GRT. Previous studies [1], [3], [4] proposed additional tests, namely, the Color Test and the Shade Test, to handle conflicts between the GRT and update transactions. However, we discovered that both algorithms can bring about nonserializable schedules of transactions. We propose an enhanced algorithm directly built on the two algorithms to guarantee the serializability of transactions. © 2006 IEEE.
KW - Concurrency control
KW - Data consistency
KW - Serializability
KW - Transaction processing
UR - http://www.scopus.com/inward/record.url?scp=33644657281&partnerID=8YFLogxK
UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-33644657281&origin=recordpage
U2 - 10.1109/TKDE.2006.1599393
DO - 10.1109/TKDE.2006.1599393
M3 - RGC 21 - Publication in refereed journal
SN - 1041-4347
VL - 18
SP - 569
EP - 572
JO - IEEE Transactions on Knowledge and Data Engineering
JF - IEEE Transactions on Knowledge and Data Engineering
IS - 4
ER -