Cloud & DevOpsAnalysis
Published per-host prices, pricing meters and Indian data regions across the 19 products in TechBag’s observability and APM guide, as of September 2026.
The 2022 version of this post asked which infrastructure monitoring tool to buy and answered with one vendor’s suite, EverestIMS, now sold as Infraon. Its pricing page still shows no number in September 2026: a free trial, a demo and a sales button, deployment on-premise, cloud or hybrid, and offices in Bengaluru and Delaware.6 So the question has to be asked differently. For an Indian IT team the decision is not which dashboard is prettiest; it is which pricing meter you can forecast, and whether the telemetry, which includes hostnames, usernames and often customer identifiers in logs, is allowed to leave the country. Both are answerable from published pages, and this piece does that.
Five products in the guide publish an infrastructure-tier price per host or node per month on annual billing. SolarWinds Observability Self-Hosted lists Essentials at US$8 a node, rising to US$14 and US$17.50 for Advanced and Premier; its SaaS sibling lists US$15.75 a node for network and infrastructure.5 Splunk Observability Cloud starts at US$15 a host for Infrastructure, US$60 with APM and US$75 end to end.4 Dynatrace lists US$29 a host for infrastructure monitoring and US$58 per 8 GiB of host memory for full-stack.3 Datadog lists Infrastructure Pro at US$15 a host a month billed annually, US$18 on demand, and Enterprise at US$23, with up to five basic hosts free; APM is US$31 a host with Infrastructure attached, US$36 on its own.2
Infrastructure monitoring lists from US$8 to US$29 a host a month, and the one India-hosted tool sits at the top
Published entry price per host or node per month, annual billing, infrastructure tier, September 2026
Source: Vendor pricing pages as recorded in TechBag's observability guide; Datadog pricing page as published at review · as of 30 September 2026 · The highlighted bar is the only product here whose vendor documents an Indian telemetry region; Datadog Enterprise is US$23 and Dynatrace full-stack US$58 per 8 GiB
The spread is 3.6 times, which sounds decisive and is not. The two cheapest lines buy different things: the US$8 SolarWinds figure is software you host and patch yourself, and Splunk’s US$15 is a starting price the final quote still moves from.1 Outside the 19, ManageEngine OpManager publishes per-device pricing with a perpetual-licence option and runs on-premises or from ManageEngine’s Chennai and Mumbai data centres; it is a network and server monitor rather than an observability platform, and for an estate of switches, servers and VMs it is often the honest answer.7
Most software a company buys is priced per user, per device or per instance, numbers that move when you hire or buy hardware. Observability is priced on how much your software is used. The guide classifies each of its 19 products by meter: 6 bill per host, 7 per event or session, 2 per gigabyte ingested, and 4 are quote-only; 15 publish a rate and 4 do not.1 Datadog alone runs three meters inside one platform: per host for APM and Infrastructure, per GB ingested plus per million events indexed for logs, and per session for real-user monitoring.1 2
Only 6 of 19 observability products bill per host; 7 bill per event and 4 publish no price
Products in TechBag's observability and APM guide by pricing meter, 30 September 2026
| Value (products) | Share | |
|---|---|---|
| Per host: Datadog, Splunk, Dynatrace, SolarWinds | 6 | 32% |
| Per event or session: Sentry, Mixpanel, two DEM products | 7 | 37% |
| Quote only | 4 | 21% |
| Per GB ingested: Datadog and Dynatrace logs | 2 | 11% |
Source: TechBag observability and APM guide, 19 product records, meter classification by the guide; TechBag count · as of 30 September 2026 · Per event includes per session and per replay; the four quote-only rows are Elastic Observability, Splunk ITSI and two Quest Foglight products
What each meter does to a forecast. Per host tracks infrastructure, so forty containers on four hosts bill as four, and the same traffic on a hundred small autoscaled instances bills as a hundred. Per gigabyte tracks logs, and the trap is that ingest and retention are often separate meters: Dynatrace prices ingest at US$0.20 a GiB, retention per GiB-day and queries per GiB scanned. Per event tracks traffic, so a campaign that works is a bill spike with no infrastructure change to explain it.1 3
The guide annotates India residency rather than eliminating on it, and the annotation is stark. Dynatrace documents an AWS Mumbai region for its three products; Elastic documents Cloud Hosted regions in Mumbai (AWS and Google Cloud) and Pune (Azure Central India). That is four rows from two vendors. The other fifteen document no Indian region: Datadog’s Asia-Pacific sites are AP1 in Japan and AP2 in Australia; SolarWinds Observability SaaS stores data in North America, Frankfurt or Australia; Splunk documents no India realm for Observability Cloud.1
Two vendors document an Indian telemetry region; the other six in the guide do not
India data-region documentation across the 19 products in TechBag's observability guide, 30 September 2026
No Indian region documented (15 rows, 6 vendors)
Indian region documented (4 rows, 2 vendors)
Source: TechBag observability and APM guide, India annotation per product; TechBag count · as of 30 September 2026 · Self-hosted products (SolarWinds Self-Hosted, Elastic self-managed) keep data wherever you run them, which is a different answer from a vendor region
Why it matters more here than in most categories: logs are where personal data hides. A web-server access log holds IP addresses, an application log holds user identifiers and sometimes email addresses, a database slow-query log holds query text with real values in it. A team that would never ship its customer table abroad ships a sampled copy of it in its logs every day. The guide’s advice is to confirm rather than assume, and Dynatrace’s own row adds the caveat that a documented region is not a written storage commitment until the contract says so.1
The original post’s buying checklist, cloud-native support, tags, custom alerts, dashboards, machine learning, one platform, is what every vendor’s datasheet now claims. None of it separates the products. The four steps below do, because they turn your estate into the numbers each meter bills on.
Four numbers from your own estate make two observability quotes comparable; a feature list never will
The inputs each pricing meter bills on, and the written answers to collect, for an Indian buyer, September 2026
Count hosts the vendor's way
Hosts, nodes, Kubernetes nodes and containers each count differently; Dynatrace counts full-stack in 8 GiB units of memory, SolarWinds counts nodes. Get the counting rule in writing.
Measure GB a day, twice
Gigabytes ingested and the share you will index or retain. Datadog meters the two apart; Dynatrace prices ingest, retention and query scans separately.
Count sessions and events
Real-user sessions, error events and replays for the per-event tools. Decide sampling before the quote, not after the first spike.
Get region, retention and INR in writing
Which region holds the data, 180 days of logs kept in India, and the price in rupees with 18% GST shown separately.
Source: Meter definitions from TechBag's observability guide; residency and retention asks from the CERT-In Directions of 28 April 2022 · as of 30 September 2026
One more option belongs on the shortlist, and it is the one the guide calls genuinely viable in this category and almost no other: Prometheus for metrics, Grafana for dashboards, OpenTelemetry for instrumentation, and Loki or Elastic for logs. It keeps every byte in your own data centre and costs an owner rather than a licence.1
Sources
Related on TechBag: the observability and APM guide holds the records this piece counts, with the narrowing tool; the Dynatrace Observability page carries the Mumbai region and the per-8 GiB arithmetic; the ManageEngine OpManager page is the on-premises, perpetual-licence route.
Disclosure. TechBag is a software reseller: it earns a margin when a reader buys through it, including from vendors named here. The analysis is TechBag’s own; no vendor paid for or saw this piece before publication. Facts are labelled reported, TechBag count or our view.