Defining the execution semantics of stream processing engines

Journal of Big Data

Table 1 Parameters identified for the SPEs under analysis

	\(\mathcal {T}\)ick	\(\mathcal {R}\)eport				\(\mathcal {S}\)cope	\(\mathcal {P}\)ick
	\(\mathcal {T}\)ick	CC	WC	NE	P	\(\mathcal {S}\)cope	\(\mathcal {P}\)ick
Time windows
Flink	Time		\(\checkmark \)	\(\checkmark \)		\(t_0 = ts_i + \beta - \omega - 1\)
Storm	Time		\(\checkmark \)	\(\checkmark \)		\(t_0 = ts_i + \beta - \omega \)
Spark	Time		\(\checkmark \)			\(t_0 = \beta - \omega - 1\)
DataFlow	Time		\(\checkmark \)	\(\checkmark \)		\(t_0 = ts_i + \beta - \omega - 1\)
Azure S.A.	Time		\(\checkmark \)	\(\checkmark \)		?
Count windows
Flink	Tuple		\(\checkmark \)	\(\checkmark \)		\(i_0 = id_i + \beta - \omega - 1\)	–
Storm	Tuple		\(\checkmark \)	\(\checkmark \)		\(i_0 = id_i + \beta - \omega - 1\)	–
Spark 1.6	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.
DataFlow	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.
Azure S.A.	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.