- BFSIBanking, Financial Services & Insurance
- IT & ITESIT services, BPM, GCCs & SaaS
- Retail & e-commerceStore chains, D2C brands & marketplaces
- HealthcareComing soon
- ManufacturingComing soon
Elastic Observability
Logs, metrics and traces on the cluster you already have — Elastic Observability is full-stack, OpenTelemetry-first observability priced on data and resources rather than per host, which changes the arithmetic once your estate is large.
Buy through TechBag
Same software. Better outcome — at a lower cost.
How it’s rated
Full scoreboard ↓Where your telemetry is stored — set by deployment
Self-managed
Stored on your infrastructure
Logs, metrics and traces stay on hardware you run. Storage location is whatever you choose.
Elastic Cloud
India storage regions documented
Hosted: Mumbai (AWS, GCP) and Pune (Azure). Serverless: Mumbai (GCP). That is a storage region — confirm processing and support access in writing.
Most observability suites are SaaS only, so a mandate to keep production telemetry in-house narrows the list quickly. Elastic stays on it because self-managed is a supported deployment, not a legacy edition. See the observability guide for how the rest of the field deploys.
Quick answer
The Elastic platform family
This page covers Elastic Observability — logs, metrics, traces and profiling. The other pillars:
Quick facts
30-second orientation- Product
- Elastic Observability (logs, metrics, APM, profiling)
- Vendor
- Elastic (NYSE: ESTC)
- Built on
- Elasticsearch — one store for every signal
- Instrumentation
- OpenTelemetry-first; EDOT collector and SDKs GA
- Deployment
- Self-managed · Cloud Hosted · Serverless
- Priced on
- Resources (Hosted) or GB ingested + retained (Serverless)
- Serverless entry
- Logs Essentials — ingest as low as $0.07/GB
- Analyst standing
- Leader, 2026 Gartner MQ for Observability Platforms
- India regions
- Documented Mumbai (AWS, GCP) and Pune (Azure) regions
- In India via
- TechBag — tier and deployment scoping, GST invoicing
Understand observability pricing before you buy it
Most product pages skip this. We start here — so you buy a capability, not a buzzword.
What is it?
Full-stack observability on Elasticsearch — logs, metrics, traces, profiles and user-experience data in one store, collected through OpenTelemetry and explored in Kibana.
A per-host SaaS monitoring stack vs Elastic Observability — the honest table
What changes when the bill follows data rather than hosts, dimension by dimension.
| Dimension | Typical per-host SaaS suite | Elastic Observability |
|---|---|---|
| Pricing axis | Per host, per container or per product | Provisioned resources, or GB ingested + retained |
| Instrumentation | Proprietary agent per vendor | OpenTelemetry via EDOT, or Elastic Agent |
| Log search | Indexed subset; rehydrate the rest | Search-native across the retention you keep |
| Deployment | SaaS only | Self-managed, Cloud Hosted or Serverless |
| Profiling | Separate add-on, code changes | eBPF Universal Profiling, no code changes |
| Retention cost | Fixed windows per plan | Hot/warm/cold/frozen, or per-GB retention |
| Honest caveat | — | Less turnkey; you design dashboards and lifecycle |
| Best fit | — | Log-heavy estates; teams already on Elasticsearch |
The data-priced, OpenTelemetry-first answer — for the most turnkey setup, weigh Datadog or Dynatrace (TechBag sells both).
The five pieces of the platform
Vendors love diagrams; buyers need to know what they’re actually operating. Here’s the whole platform, demystified.
EDOT & Elastic Agent
Elastic Distributions of OpenTelemetry — collector plus language SDKs — or Elastic Agent. OTel data lands without a translation layer, so instrumentation stays portable.
Streams
Streams routes raw logs into logical streams and uses AI to extract structure from them, so you stop hand-writing parsing pipelines for every new source.
Elasticsearch
Logs, metrics, traces and profiles in one engine. Elastic says it stores logs and traces up to 4x more efficiently than standard indexing, and metrics up to 2.5x better than Prometheus.
Kibana apps
APM service maps, infrastructure inventory, log explorer, synthetics and RUM views, SLOs and alerting — one UI over the same data rather than a tab per tool.
ML & AI Assistant
Anomaly-detection jobs and log categorisation find the odd pattern; the AI Assistant helps explain an error or a trace in plain language during an incident.
OpenTelemetry in, one Elasticsearch store underneath, Kibana on top — every signal from an incident in the same place.
Twelve capabilities. Ingest, analyse, act.
Elastic Observability collects logs, metrics, traces and profiles through OpenTelemetry, analyses them in one store, and turns them into SLOs, alerts and answers during an incident.
EDOT collector and SDKs
Elastic’s OTel distributions — collector plus Java, .NET, Node.js, Python, PHP and mobile SDKs — all GA.
Log analytics at volume
Collect, search and explore large log volumes, with Streams extracting structure from raw lines.
Prometheus-native metrics
Scrape or remote-write Prometheus metrics alongside host and container metrics, in the same store.
Tiered data retention
Hot, warm, cold and frozen tiers on self-managed and Hosted; per-GB retention pricing on Serverless.
Distributed tracing
Service maps, transaction breakdowns and trace waterfalls that link a slow request to its logs.
Hosts, Kubernetes and cloud
Inventory and health views for on-premises and cloud infrastructure, down to the pod.
Universal Profiling
eBPF whole-system profiling with no code changes — Elastic cites under 1% CPU overhead.
Synthetics, RUM and uptime
Scripted browser tests, lightweight checks and real-user monitoring against the same backend data.
Anomaly detection & log patterns
Machine-learning jobs baseline normal behaviour and group noisy log lines into a handful of patterns.
SLOs and alerting
Define service-level objectives and burn-rate alerts on the signals you already collect.
Elastic AI Assistant
Ask what an error means or what changed before a spike, grounded in the data in your cluster.
Workflows and Agent Builder
Newer building blocks for automated follow-up and custom AI agents — metered per execution on Serverless.
Watch Elastic Observability in action
The overview, APM tracing on a real application, and a first deployment.
Elastic Observability Overview
The observability solution, explained by Elastic.
Elastic Observability APM: Observing an application
Tracing a real application end to end.
Getting Started with Elastic Observability
What the first deployment looks like.
Want a live, India-context walkthrough for your environment?
Book a guided demo →Pay for data, not for every host you run.
Here’s what genuinely sets Elastic Observability apart (and where Datadog, Dynatrace or Grafana may fit better).
Priced on data and resources, not per host
Hosted bills the cluster you provision; Serverless bills GB ingested and retained. Neither counts hosts or containers, so a Kubernetes estate that autoscales does not multiply the bill by node count. Model your own volumes before assuming it is cheaper.
OpenTelemetry-first, so you can leave
EDOT is Elastic’s own build of the OpenTelemetry collector and SDKs, generally available, with fixes contributed upstream. Instrument once in OTel and the data is portable — a hedge most proprietary agents do not give you.
Search is the native operation
Logs are the biggest, messiest signal, and Elastic is built on a search engine. Ad-hoc questions across months of logs are what the platform does by default, not a premium query tier bolted onto a metrics store.
Self-managed is a real option
Most observability suites are SaaS only. Elastic runs on your own hardware, on Hosted in 50+ regions, or Serverless — so a mandate to keep telemetry on your own infrastructure does not remove it from the shortlist.
The honest caveat — you tune it
Out of the box, Datadog and Dynatrace do more for you: auto-instrumentation, curated dashboards, fewer knobs. On Elastic you design index lifecycle, shard sizing and dashboards, unless Serverless does it. Price that time.
The honest positioning
Right when log volume dominates, when you already run Elasticsearch, or when OTel portability and self-managed options matter. Wrong when you want the most turnkey SaaS and will pay per host for it. TechBag sells the field and says which fits.
The numbers behind the platform
What your Elastic Observability rollout looks like
Find the cluster you already run
Many teams already ship logs to Elasticsearch. Start from it, and decide what APM and metrics add. TechBag scopes this free.
Pick deployment and pricing axis
Self-managed, Hosted (resources) or Serverless (per GB). This choice sets the bill and where the data is stored.
Instrument with OpenTelemetry
Roll out EDOT to a few services first, set retention tiers before volume arrives, then widen to the estate.
SLOs, alerts and profiling
Define SLOs, tune alert noise, turn on profiling where CPU cost hurts. TechBag supports renewals in INR with GST.
The review scoreboard
Modelled on Gartner Peer Insights structure. *Counts and breakdowns are illustrative pending verified review collection.
Quick poll — what’s driving your evaluation?
Talk to an advisor“Our per-host APM bill tracked every autoscaling event. Moving traces to Elastic turned it into a storage question we could actually plan for.”
“We instrumented with the EDOT Java agent. If we ever move, the OpenTelemetry work comes with us — that was the deciding point for our architects.”
“Log search is where it earns its keep. Grepping six weeks of payment-gateway logs during an incident takes seconds, not a ticket to another team.”
“Dashboards took real effort. Datadog would have shipped more out of the box; here we built our own views before the team trusted them.”
“Universal Profiling found a JSON serialiser eating a fifth of our CPU on one service. We had never instrumented it, and did not need to.”
“Set retention tiers before the first big ingest. We learnt that after three months of hot storage we did not need.”
“Streams saved us writing grok patterns for a dozen legacy apps. Not perfect on every format, but most lines came out structured.”
“Our regulator wanted production telemetry kept in-house. Self-managed Elastic let us keep APM and logs on our own racks.”
Where everyone sits — the grids
Analyst firms bury this view behind paywalls, and review-site grids score satisfaction, not fit for your estate. So here’s TechBag’s synthesis of the observability-platform market — tap any vendor to see why it sits where it does.
TechBag Observability Platforms Grid
Execution strength vs product vision — the classic market map, minus the paywall.
Leader in the 2026 Gartner MQ for Observability Platforms, third year running.
Openness × Cost Predictability
The grid nobody publishes — OpenTelemetry and self-host freedom vs how predictable the bill stays as the estate grows.
OTel-first, self-managed option, priced on data not hosts.
Positions are TechBag’s illustrative synthesis of public review-platform data and vendor documentation — not a reproduction of any analyst graphic. Verify before relying on it.
Elastic Observability vs the field
The observability field — honest lanes; Elastic’s edge is search-native logs, OpenTelemetry portability and pricing on data rather than hosts. Most turnkey setup? Datadog or Dynatrace. We say so (and sell them).
| Dimension | Elastic Observability | Datadog | Dynatrace | Grafana (LGTM / Cloud) | New Relic | Splunk Observability |
|---|---|---|---|---|---|---|
| Position | Observability on a search engine | Broad SaaS monitoring platform | Auto-instrumented, AI-led platform | Open-source stack + managed cloud | SaaS all-in-one observability | SaaS APM + infrastructure |
| Deployment | Self-managed, Hosted or Serverless | SaaS only | SaaS or self-hosted Managed | Self-host OSS/Enterprise or Cloud | SaaS only | SaaS; Splunk Enterprise for on-prem logs |
| Pricing axis | Resources, or GB ingested + retained | Per host plus per-GB logs, per product | Consumption platform subscription | Usage-based; free tier; OSS is free | GB ingested plus user seats | Largely host-based |
| OpenTelemetry | OTel-first; EDOT distributions GA | Supported beside its own agent | Supported beside OneAgent | OTel-native, own collector (Alloy) | OTel-native ingest | Built on OTel collector |
| Log search depth | Search engine is the core | Strong; indexing drives cost | Grail lakehouse | Loki — label-indexed, cheaper, less ad hoc | Solid log management | Splunk platform for logs |
| Out-of-the-box setup | More configuration; Serverless eases it | Very turnkey, huge integration list | OneAgent auto-discovery | Assemble and tune the stack | Quick start, guided installs | Moderate |
| Continuous profiling | eBPF Universal Profiling | Profiler, priced separately | Code-level via OneAgent | Pyroscope | Available | AlwaysOn Profiling |
| Analyst standing | Leader, 2026 Gartner MQ | Check the 2026 report | Check the 2026 report | Check the 2026 report | Check the 2026 report | Check the 2026 report |
| Operational burden | Yours if self-managed; none on Serverless | Fully managed | Managed (SaaS) | Yours if OSS; managed on Cloud | Fully managed | Managed (SaaS) |
| Best fit | Log-heavy estates, OTel, self-host options | Turnkey breadth, budget for per-host | Large enterprises wanting automation | Open-source-first teams | Simple data + user pricing | Estates already on Splunk |
Does Elastic Observability fit you?
Honest fit signals — because the fastest way to lose your trust is to pretend one product wins every scenario.
Choose Elastic Observability if…
- ✓Logs are your biggest signal and ad-hoc search across them matters
- ✓You already run Elasticsearch, so adoption is enabling rather than migrating
- ✓You want OpenTelemetry portability and a self-managed option
- ✓Per-host pricing is what hurts in an autoscaling estate
Choose Datadog if…
- ✓You want the most turnkey SaaS with the widest integration list, and the per-host bill is acceptable
Choose Dynatrace if…
- ✓You want OneAgent auto-discovery and AI-led root cause across a large enterprise estate
Choose Grafana if…
- ✓Your team is open-source-first and happy to assemble Loki, Mimir and Tempo — or use Grafana Cloud
Choose New Relic or Splunk if…
- ✓You want simple data-plus-users pricing (New Relic), or your logs already live in Splunk
Elastic Observability is one of 23 observability & APM products TechBag carries. The Observability & APM guide narrows them to a shortlist and shows the reasoning. →
Self-managed, or managed?
Self-managed and Cloud Hosted Elastic Observability are priced on subscription tier and resources, not per host — so the question that decides your bill is whether running the observability cluster yourself costs less than letting Elastic run it. The engine is free; what is not free is your engineers’ time on shard sizing, index lifecycle, capacity and upgrades. (Serverless, billed per GB, removes the cluster work entirely.) Drag both sliders. Illustrative, at ₹12,000 per fully-loaded engineer-day and ~₹84/USD.
Engineer-days means real time on shard strategy, index lifecycle, capacity planning and version upgrades — not firefighting a production outage. If you put 0, you are assuming the cluster runs itself, which is the assumption this calculator exists to question. Illustrative: your TechBag quote models your real tier and resources.
Three ways to pay for it
Elastic Observability is never priced per host. Self-managed pays a subscription tier on hardware you run; Cloud Hosted bills the resources you provision; Serverless bills the GB you ingest and retain. Figures below are Elastic’s published “as low as” list rates in USD and scale with your volumes. TechBag scopes tier and deployment and quotes in INR with GST.
Self-managed
Best if telemetry must stay in-house
- Engine free to run — paid tiers add features and support
- Data stored on infrastructure you control
- You operate the cluster — price your engineers’ time
Elastic Cloud Hosted
Best for control without running it
- From $99/mo Standard, $114 Gold, $131 Platinum, $184 Enterprise
- Reference size: 120 GB storage across 2 zones; scales with resources
- 50+ regions across AWS, GCP and Azure
Serverless Observability
Best for no cluster to run
- Logs Essentials: ingest from $0.07/GB, retention from $0.017/GB-month
- Complete: logs and traces from $0.09/GB, retention $0.019/GB-month
- Egress: 50 GB free, then $0.05/GB; synthetics billed per run
Buy it for less — TechBag pricing beats list
Whatever the list prices above, TechBag negotiates a significantly better deal — with GST-compliant INR invoicing and local support. Ask us for your discounted quote.
Get an India-ready quote
Tell us your requirements and current tools — we’ll model it against what you spend today.
The 8 questions to ask every observability vendor
Take this into your next vendor call — including ours.
Is the bill driven by hosts, containers, GB ingested or provisioned resources? Model your own estate on each vendor’s axis.
Will you instrument with OpenTelemetry, or a proprietary agent? What does leaving cost you in re-instrumentation?
How many GB a day of logs, and how long must they stay searchable? That, not host count, drives most bills.
Which data needs hot search, and which can sit in cheaper tiers? Set the policy before the first big ingest.
SaaS, self-managed, or both? If telemetry must stay on your infrastructure, which vendors remain on the list?
Who designs dashboards, index lifecycle and alert rules? Have you priced their time alongside the licence?
Are continuous profiling, synthetics and real-user monitoring included, or separately priced add-ons?
Is the quote in USD? Have you modelled it in INR with GST, including egress and synthetic test runs?
Questions buyers ask
Ready to evaluate Elastic Observability?
Model your log and trace volumes on Hosted, Serverless or self-managed, or get an honest Elastic-vs-Datadog comparison — in INR/GST.
Stats, ratings, review counts and pricing are illustrative and sourced from public materials; verify before purchase.