Monitoring aplikacija – Važnost preventive
IBM Instana
Monitoring versus observability
IBM Instana® Observability automatski otkriva, mapira i nadgleda sve servise i komponente infrastrukture, kako u lokalnom (on-prem) okruženju, tako i u cloud-u, pružajući kontekst aplikacija zasnovan na veštačkoj inteligenciji i mogućnosti brze detekcije i rešavanja problema, čime se unapređuje rad IT operacija. Instana-ine nadzorne table bez potrebe za konfiguracijom omogućavaju trenutno otklanjanje grešaka i praćenje sistema, dok automatizovani playbook-ovi rešavaju uobičajene probleme, a precizna upozorenja zasnovana na mašinskom učenju pomažu u upravljanju brzim promenama.
Za kompanije u Srbiji koje žele da unaprede nadzor svojih aplikacija i infrastrukture, tu je COMING – pouzdan IT partner i zvanični IBM Silver Partner. Naša kompanija nudi lokalnu stručnost, tehničku podršku i implementaciju IBM Instana rešenja, uz razumevanje specifičnih potreba domaćih organizacija. Bilo da tek ulazite u svet observability-ja ili želite da automatizujete i modernizujete postojeći nadzor, COMING vam može pomoći da izgradite efikasno i skalabilno rešenje.
What is monitoring?
Monitoring is the process of using pre-configured telemetry data with dashboards and alerts to understand your application’s health and performance.
Outputs
monitoring
examines
Health checks
Health checks
periodically poll a
specific service for
a success response.
Alerts
Alerts are triggered
by specific predefined
thresholds being
exceeded.
Dashboards
Dashboards show specific predetermined metrics.
Traces
Traces show how
operations move
from one node to
another throughout
your systems.
Logs
Logs are a timestamp
record of events
happening in your
software.
Metrics
Metrics use
time-series data
to monitor system
performance
%
of executives surveyed
said mainframe-based
applications are
central to their
business strategy.1
What is
observability?
Observability is the ability to understand the inner
state of your evolving systems by analyzing all
available outputs in real time.
Outputs
observability
examines
Distributed traces
Distributed traces
follow an entire
request lifecycle
through all services.
Automated service discovery
New services and
applications can
be discovered
and observed
without additional
deployments or
configuration.
Dynamic dependency graph
Communication
between services is
mapped and graphed
as it occurs.
Contextualized logs
Logs can be correlated
with specific traces,
metrics, services
and hosts for faster
debugging.
Granular metrics without sampling
Unsampled metrics
allow for complete
awareness and faster
responses.
Application profiling
Technology-specific
instrumentation
gathers metrics
from within service
runtimes.
End user monitoring
Instrumentation in
mobile apps and web
front ends allows for
monitoring real user
experiences.
Smart alerts
Smart alerts can
be triggered by
dynamically defined
criteria based on
previous metrics.
Root cause analysis
AI-powered expert
knowledge ingests all
of this observability
data to track down the
root causes of errors.
Relationship between
observability and monitoring
Monitoring
– measures system health by collecting and
analyzing aggregate data systems using
predefined metrics and logs.
– helps teams detect known failures reactively,
with some limitations.
– requires you to know and specify which
metrics and logs to track.
Observability
– helps you understand a complex system’s
internal state based on external outputs.
– allows teams to identify performance issues
using system data.
– allows proactive prevention and requires no
additional testing or coding.
1. IBM Institute for Business Value: https://www.ibm.com/thought-leadership/institute-
business-value/report/application-modernization-mainframe
© Copyright IBM Corporation 2023. IBM, the IBM logo, and IBM Instana® are trademarks or
registered trademarks of International Business Machines Corporation, in the United States
and/or other countries. Other product and service names might be trademarks of IBM or other
companies. A current list of IBM trademarks is available on ibm.com/trademark.