OpenTelemetry for Mainframe Observability
Key Points
- Modern hybrid applications span front‑ends, cloud services, and mainframes, so end‑to‑end visibility is essential for reliable operation.
- Most organizations rely on four to seven separate monitoring tools, creating a fragmented stack that slows detection, isolation, and resolution of issues.
- OpenTelemetry, a vendor‑agnostic, CNCF‑governed framework, standardizes the generation, processing, and transmission of metrics, traces, and logs across diverse environments.
- By instrumenting code (automatically or via OpenTelemetry SDKs) and using the OpenTelemetry Protocol, spans are collected and processed by the OpenTelemetry Collector, which can filter, enrich, and forward data to downstream systems.
- OpenTelemetry itself is not a product; it integrates with a variety of observability back‑ends and tooling to provide a unified view of system behavior.
Sections
Full Transcript
# OpenTelemetry for Mainframe Observability **Source:** [https://www.youtube.com/watch?v=hLvwoow3XTk](https://www.youtube.com/watch?v=hLvwoow3XTk) **Duration:** 00:05:32 ## Summary - Modern hybrid applications span front‑ends, cloud services, and mainframes, so end‑to‑end visibility is essential for reliable operation. - Most organizations rely on four to seven separate monitoring tools, creating a fragmented stack that slows detection, isolation, and resolution of issues. - OpenTelemetry, a vendor‑agnostic, CNCF‑governed framework, standardizes the generation, processing, and transmission of metrics, traces, and logs across diverse environments. - By instrumenting code (automatically or via OpenTelemetry SDKs) and using the OpenTelemetry Protocol, spans are collected and processed by the OpenTelemetry Collector, which can filter, enrich, and forward data to downstream systems. - OpenTelemetry itself is not a product; it integrates with a variety of observability back‑ends and tooling to provide a unified view of system behavior. ## Sections - [00:00:00](https://www.youtube.com/watch?v=hLvwoow3XTk&t=0s) **Untitled Section** - ## Full Transcript
modern applications do not exist in
silos or monolithic technology Stacks
typically today we see front-end
applications triggered perhaps by mobile
devices connecting to cloud services and
then onto other Technologies including
the Mainframe in this video I'm going to
be talking about observability open
Telemetry and how it can relate to the
Mainframe in these applications the
transactions and interactions between
the systems are critical for it
operations to understand endtoend
visibility is key to knowing whe the
hybrid application is performing
correctly most organizations use a
number of monitoring and observability
tools to achieve this however we know
that most organizations actually use
between four and seven different Tools
in order to do this this disjointed
approach means that we end up with uh a
slower process to actually detect
isolate and ultimately resolve it
operational issues to address this
challenge we've seen the emergence of
open Telemetry over the last 5 years so
what is it it we'll put simply open
Telemetry is a vendor agnostic
observability framework that helps with
the generation processing and
transmission of different types of
telemetry data or signals these are
metrics
traces and
logs now what makes open telemetry
different is that unlike a vendor
specific
implementation open Telemetry is an open
framework that is managed by the cloud
native Computing foundation and has
support and contributions from many
different vendors across the IT industry
now we're going to talk about some of
the basic concepts of open Telemetry
first your application or the
environment it's running on needs to be
instrumented this will allow the signals
to be generated in order to observe the
system
this can either be done automatically
depending on the language that you're
using or it might have to be done
manually using sdks provided by open
Telemetry the system will then generate
data in the form of open Telemetry
protocol or
OTP this describes properties of the
system that are being observed and
seen by complying with this standard
different spans from different
Technologies and different systems can
then be later correlated together to
give the full observability view of the
system these spans are sent to the open
Telemetry
collector The Collector is a process
that takes this information and performs
processing on it such as filtering or
maybe enriching the data in some form by
perhaps correlating multiple sources
together it can then be sent on to other
collectors or maybe an observability
back end for visualization and further
analysis so as you instrument all of
these across all these environments we
will now generate the data that allows
the full system to be observed so now we
have all of this Telemetry data and
probably quite a lot of it what do we do
with it this is where the observability
back end comes
in now to be clear open Telemetry itself
is not a product
instead there are many different
software offerings out there both open
source and vendor provided that take OTP
Source data process it and perform
analysis on it this will take data from
multiple from all these sources derive
visualizations and Analysis around
performance and the precise nature of
this analysis varies from product to
product but typically they will provide
an analysis of the application that's
running and the infrastructure that that
it that the application is running on an
obvious example of this would be a trace
view of a critical customer Journey from
a mobile initiated transaction through
the environment to maybe a database
update on the
Mainframe using open Telemetry we can
then enrich that information using
signals from metrics and logs across all
these different environments bringing
them together to give a more complete
view this helps the it operations teams
and site reliability Engineers to more
quickly identify where where a problem
might be if there is a performance issue
isolate it to the particular area that
is impacting it and then taking prompt
action whether that's something they do
themselves or engaging the right subject
matter expert to then take action on
that quickly in this video we've learned
the basics about what open Telemetry is
but we just scratched the surface of
what is possible by exploiting it
observability provides deep
visualization into hybrid applications
that span multiple technology stacks and
provide automated res solutions to
problems within there having a standard
adopted across all these platforms means
that it is easier to integrate that data
and build up that full complete view
this is true for not just new
applications that are written on the
cloud in microservices or containers but
also older applications that might have
been around for many years running on
something like the Mainframe when you
have a complete endtoend view the time
and resources taken to resolve a problem
are greatly
reduced e