3. Computação Ubíqua
“The most profound technologies are those that
disappear. They weave themselves into the
fabric of everyday life until they are
indistinguishable from it.”
(Mark Weiser – The Computer for the 21st Century)
16. 16slide 16
Internet of Things
Fonte: http://www.123rf.com
Desafios:
Heterogeneidade de dispositivos
Escalabilidade
Troca de dados ubíqua
Otimização de energia
Localização e rastreamento
Autogerenciamento
Gerenciamento de dados
Semântica de interoperabilidade
Segurança e privacidade
Mobilidade
26. 26slide 26
Como processar esses dados?
2004
MapReduce é um modelo de programação, e framework introduzido pelo
Google para suportar computações paralelas em grandes coleções de
dados em clusters de computadores
30. 30slide 30
Map Reduce
• But MR is Batch Processing…
– V = Volume
– Time is not an issue
– Has access to all data
– Might compute something big and complex
– Is generally more concerned with throughput than
latency of individual components of the computation
– Has latency measured in seconds/minutes
• Depends on the size of the data
38. 38slide 38
Cenários
O que esses cenários tem em comum?
Fluxo contínuo de dados
Como você projetaria/implementaria um sistema que
fica monitorando dados bancários, checando dados
do twitter e variação da bolsa de valores?
54. CEP (Complex Event Processing)
• Based on publish-subscribe
– Topic vs. content-based
– subscriptions may refer to single events only and
cannot take into account the history of already received
events or relationships between events
• CEP as an extension of publish-subscribe
(register for composite events)
• detect complex patterns of incoming items, involving
sequencing, ordering relationships creating patterns
(different from DSMS)
55. CEP (Functional Model)
Internal memory
not present in all IFP
engines
(Smoke, temperature) Fire Alert (Notify chemical alert)
57. EPL (Rules)
• The Event Processing Language (EPL) is a SQL-
standard language with extensions
– offering SELECT, FROM, WHERE, GROUP
BY, HAVING and ORDER BY clauses.
– Streams Tables
– Events Rows
• Since events are composed of data
– joins, filtering and aggregation through grouping
64. Twitter Heron
Twitter has completely replaced Storm with Heron which is
currently processing “several tens of terabytes of data,
generating billions of output tuples” per day, “delivering 6-14X
improvements in throughput, and 5-10X reductions in tuple
latencies” on a standard word count test, and resulting in a 3X
reduction in hardware