An Overview of the NoSQL World - Large Scale and Big Data: Processing and Management - page 312

their business goals. The state-of-the-art cloud databases do not allow the speci-
fication of SLA metrics at the application nor at the end-user level. In practice,
cloud service providers guarantee only the availability (uptime guarantees), but
not the performance, of their services [6,10,31]. In addition, sometimes the granu-
larity of the uptime guarantees is also weak. For example, the uptime guarantees
of Amazon EC2 is on a per data center basis where a data center is considered to
be unavailable if a customer cannot access any of its instances or cannot launch
replacement instances for a contiguous interval of five minutes. In practice, tra-
ditional cloud monitoring technologies (e.g., Amazon CloudWatch ) focus on low-
level computing resources (e.g., CPU speed , CPU utilization , I/O disk speed ). In
general, translating the SLO of software application to the thresholds of utilization
for low-level computing resources is a very challenging task and is usually done
in an ad hoc manner due to the complexity and dynamism inherent in the interac-
tion between the different tiers and components of the system. Furthermore, cloud
service providers do not automatically detect SLA violation and leave the burden
of providing the violation proof on the customer [10].
In the multi-tenancy environment of DaaS, it is an important goal for DaaS
providers to promise high performance to their tenants. However, this goal nor-
mally conflicts with another goal of minimizing the overall running servers and thus
operating costs by tenant consolidation. In general, increasing the degree of multi-
tenancy (number of tenants per server) is normally expected to decrease per-tenant-
allocated resources and thus performance, but on the other hand, it also reduces the
overall operating cost for the DaaS provider and vice versa. Therefore, it is neces-
sary, but challenging for the DaaS providers to balance between the performance
that they can deliver to their tenants and the data center's operating costs. Several
provider-centric approaches have been proposed to tackle this challenge. Chi et al.
[22] have proposed a cost-aware query scheduling algorithm, called iCBS , that takes
the query costs derived from the SLAs between the service provider and its custom-
ers (in terms of response time) into account to make cost-aware scheduling decisions
that aims to minimize the total expected cost. SLA-tree is another approach that
have been proposed to efficiently support profit-oriented decision making of query
scheduling. SLA-tree uses the information about the buffered queries that are wait-
ing to be executed in addition to the SLA for each query that indicates the differ-
ent profits for the query for varying query response times and provides support for
the answering of certain profit-oriented what if type of questions. Lang et al. [46]
presented a framework that takes as input the tenant workloads, their performance
SLA, and the server hardware that is available to the DaaS provider, and produces
server characterizing models that can be used to provide constraints into an opti-
mization module. By solving this optimization problem, the framework provides a
cost-effective hardware provisioning policy and a tenant scheduling policy on each
hardware resource. The main limitation of this approach is that the input information
of the tenant workloads is not always easy to specify and model accurately. PIQL
( P erformance I nsightful Q uery L language [5] is a declarative language that has been
proposed with a SLA compliance prediction model. The PIQL query compiler uses
static analysis to select only query plans where it can calculate the number of opera-
tions to be performed at every step in their execution. In particular, PIQL extends