Which CMDB fits change and impact management better than asset tracking?
If your team needs a CMDB to answer 'what breaks if this changes?', iTop is the stronger fit. Its fully customizable CMDB data model, change management and graphical impact analysis are built to map services, contracts and configuration items to each other and to show the downstream effect of a change. GLPI is the stronger fit when the core job is knowing what hardware and software you own, where it sits and what it is licensed for, because its asset-centric inventory is fed by the GLPI Agent. The deciding question is how much relationship modeling your team needs and who will keep that model accurate over time.
Modeling depth is the first test. If you need custom classes, relationships between services and components, and SLA and contract logic tied to those relationships, iTop's data model gives you that flexibility, but you will design and maintain it. If you mainly need reliable discovery of endpoints, installed software and asset records, GLPI's native inventory with the GLPI Agent covers most of that with less modeling work. Network discovery and SNMP inventory need the separate GLPI Inventory plugin, so confirm that scope before you assume it is included. Ask each vendor or partner to demonstrate an impact analysis on one of your real change requests, using your own naming and relationships.
Ownership is the second test. A relationship model decays quickly if no one is accountable for updating it after each change, so assign data owners and a review cadence before you pick the platform. iTop ships data consistency audits and CSV import to help keep records clean, which helps only if someone runs them. Plan for the operating cost of each option: upgrade effort, plugin compatibility and the time your administrators spend on data quality. Check the current release notes and supported database versions in the official iTop software requirements before you size the server.
How GLPI and iTop split the CMDB work
This table compares how each platform handles the CMDB, discovery, change and impact work, and the licensing and vendor support behind them. Read it by asking which gaps matter for your team, then confirm any capability marked as needing a plugin or confirmation before you commit.
| Capability | GLPI | iTop |
|---|---|---|
| Data model | Asset-centric inventory with ITIL tickets, problems and changes attached to assets | Fully customizable CMDB data model, described in the iTop README as a core feature |
| Native inventory | GLPI Agent collects hardware and software inventory natively since version 10 | Inventory is not a built-in agent feature in the supplied facts; confirm agent or connector options with Combodo |
| Network discovery | Needs the separate GLPI Inventory plugin for network discovery and SNMP inventory | Confirm discovery options and any extensions in the iTop Hub for your environment |
| Change management | ITIL change records linked to assets and tickets | Change management is listed in the iTop README alongside customizable SLA management |
| Impact analysis | Asset and ticket relationships; confirm how far graphical impact views go for your model | Graphical impact analysis is listed in the iTop README |
| CSV import and consistency audits | Confirm CSV import and data validation workflows against the current documentation | CSV import and data consistency audits are listed in the iTop README |
| Licensing | GPL-3.0-or-later, maintained by Teclib' | AGPL-3.0, sponsored, led and supported by Combodo |
| Paid editions | Teclib' GLPI Network subscriptions, priced per user or by tier for self-hosted use | Community free up to 10 agents; Atelier up to 50 agents; Forge above 50, on premises or SaaS |
| Vendor support | Teclib' sells GLPI Network subscriptions and cloud hosting | Combodo sells commercial support, training and consultancy; confirm scope in your quote |
How the data model shapes change and impact work
The data model decides how far impact analysis can go, and this is where iTop and GLPI diverge most. iTop's README describes a fully customizable CMDB with graphical impact analysis, change management, customizable SLA management and data synchronization for federation. That means you define the object classes, their attributes and the relationships between them, and the impact view can only trace what you have modeled. Before you rely on it for change approval, list the business services, the application components, the servers and network devices they run on, and the contracts and teams that own each one, then decide which relationship types (depends on, hosted on, used by) you will maintain as data.
GLPI approaches the CMDB differently. Its assets, contracts, ITIL tickets, problems and changes sit in one suite, and the GLPI Agent feeds hardware and software inventory into the asset records. That makes GLPI strong at answering what is installed where and who holds the contract, but the depth of service-to-component relationships is a question to check before you commit. Read the GLPI assets documentation and confirm whether the item types, relations and graphical views you need for impact tracing exist in the version you plan to run, and whether any of them depend on plugins.
In a pilot, pick one change that touched a shared service and ask both candidates to show the affected components, the owning teams and the related open changes. Time how long a data engineer needs to add a new class or relation in each product, and record who will maintain that model after go-live. The model you cannot keep current is the one that will give you wrong impact answers, so budget for that upkeep in whichever option you choose.
What changes the result in practice
Four to six factors decide whether GLPI or iTop fits a CMDB-driven team. Most of them come down to how your CI data gets created, how much model you need to customize, and what your team can run and extend.
Custom classes and relations
iTop's data model is fully customizable, so you can define your own classes, attributes and relations between configuration items and services. That matters when impact analysis must follow your specific dependencies. GLPI's asset model is organized around its built-in asset types, so check whether the types and relations you need already exist before you commit. Ask each vendor to model two or three of your real services during evaluation.
Endpoint inventory or a maintained model
If hardware and software must arrive automatically from endpoints, GLPI's native inventory through the GLPI Agent is the more direct path, since the agent runs on Windows, macOS and Linux and reports to GLPI. If your CMDB is curated by hand from change records and service owners, iTop's model suits that workflow. Decide which source is authoritative for each CI type before you pick the tool, because mixing sources without rules creates duplicate records.
Discovery and SNMP needs
Network discovery and SNMP inventory in GLPI require the separate GLPI Inventory plugin, so it is an additional component to install, configure and maintain. Confirm whether your network devices, switches and appliances must be discovered automatically, and whether that discovery feeds the same CI records your change process uses. Treat the plugin as part of the GLPI scope in any quote.
Change approval workflows
Impact analysis is only as useful as the change process around it. iTop covers change management and graphical impact analysis in its core product, which suits teams that must trace a proposed change through dependent services before approval. GLPI handles ITIL changes and problems alongside assets, which suits teams whose change needs are lighter. Map your approval steps, required approvers and rollback records against each product in a pilot.
Agent counts and edition caps
Combodo positions iTop Community for up to 10 agents, Atelier for up to 50 and Forge for more than 50, in SaaS or on premises. GLPI Network tiers are set by agent and asset caps, with self-hosted Basic covering up to 10 agents and 500 assets. Count your managed endpoints, users and assets now and project growth, because the tier you pick sets the license and support cost for both products.
PHP and database skills
iTop 3.3.x needs PHP 8.2 to 8.4, MySQL 5.7 or later or MariaDB 10.6 to 11.8, with MariaDB preferred, and GraphViz. GLPI 12.0.0 needs PHP 8.3 or later with MariaDB 10.11 or later or MySQL 8.0 or later. Check whether your team can maintain these stacks, patch them and extend the data model, and whether a partner will handle that work under a fixed scope.
Which platform fits which team
Teams whose main question is what hardware and software they own, where it sits and who holds it will usually find GLPI the more direct fit. GLPI pairs ITIL ticketing with asset management, and it has native inventory through the GLPI Agent since version 10, which covers hardware and software collection on Windows, macOS and Linux. Network discovery and SNMP inventory need the separate GLPI Inventory plugin, so plan for that addition if switches, printers or network gear are in scope.
- Endpoint inventory: The GLPI Agent sends hardware and software data into GLPI without a separate inventory server to run.
- Discovery scope: Confirm whether network discovery and SNMP work is required, since it depends on the GLPI Inventory plugin.
- Contract and license tracking: GLPI's asset model is built for ownership, contracts and financial records alongside tickets.
- Ongoing agent work: Agents must be deployed and kept current across the estate, which is a recurring operational task.
- Change depth: If change approval is secondary, iTop's extra CMDB modeling may add effort without adding value.
Before committing, ask Teclib' and Combodo to confirm which discovery and inventory features each edition supports in your environment.
Teams that must show which services depend on which components before approving a change will usually find iTop the closer fit. iTop describes a fully customizable CMDB with graphical impact analysis, change management and customizable SLA management, so the data model can follow the way your services are actually built. This matters most when a change to one database or network segment has to be traced to the applications and contracts it affects.
- Service modeling: Customize classes and relationships to match how your applications, middleware and infrastructure depend on each other.
- Impact analysis: iTop's graphical impact view supports approval decisions where dependency traces are the main input.
- Data quality: Data consistency audits and CSV import help keep the model usable as it grows.
- Federation: Data synchronization is available when the CMDB must feed or receive data from other instances.
- Modeling effort: A richer model needs design work up front, and that effort is usually led by the team or by an implementation partner.
Ask Combodo or a partner to walk through how a sample change would be approved in iTop, and confirm the edition and agent count that fits your estate.
Some organizations need an asset register that the service desk trusts and a dependency model that change managers can use, and they are not willing to maintain two sources of truth. In that case the decision is less about which product is better and more about where the system of record lives and which tool owns each data type. Either platform can be central, but the choice has to be made before data is migrated.
- Single source of truth: Decide which system owns hardware records and which owns service relationships, then keep the link explicit.
- Synchronization: Confirm whether an integration or scheduled import is needed, and who maintains it after go-live.
- Edition boundaries: iTop's agent-tier editions and GLPI's subscription levels both affect cost once the estate grows, so model both before choosing.
- Team ownership: Name a data owner for each record class so stale asset or dependency data does not reach approvals.
- Parallel running: Running two tools long term doubles administration, patching and backup work.
Before a parallel design is approved, ask both vendors and your implementation partner to document the integration path, the upgrade process for each platform and the support model for the combined setup.
How to test CMDB depth in your own environment
This sequence gives you a like-for-like test of how each platform models your assets, relationships and changes, so you can compare CMDB depth on your own data rather than on vendor claims. Each step produces a record you can compare across GLPI and iTop.
Write the impact questions first
Before touching either platform, the architecture or service owner lists the impact questions the CMDB must answer, such as which business services depend on a given database host, which tickets and changes touch a server, and which owners must approve a maintenance window. Keep the list to questions your change advisory board actually asks. Done means every question has a named owner and a written expected answer, so each platform can be judged against the same list.
Export a representative asset sample
The CMDB or inventory owner exports a sample of real configuration items and their relations from your current source, usually a CSV export from an existing CMDB, spreadsheet or discovery tool. Include servers, applications, databases, network devices and their relationships, and remove fields that contain sensitive data. Done means the file has a documented column mapping and a record count the team agrees represents the mix of asset types in production.
Load the sample into each platform
An administrator imports the same CSV into a test instance of GLPI and a test instance of iTop, using each platform's import path and a data model that reflects how you want assets and relations to be represented. Record every field that needed a transformation, every relation that could not be expressed directly, and every row rejected on import. Done means both instances hold the sample and you have a written list of mapping gaps for each platform.
Model one change with dependent services
Pick one realistic planned change that touches a shared component, such as a database upgrade, and record its affected configuration items, dependent services and approvers in each platform. Use the change workflow each product offers, and note which relationships you could follow natively and which required custom fields, plugins or manual steps. Done means each platform shows the same impact path for that change, or the gap is documented with the workaround used.
Run the data consistency checks
Run each platform's own integrity checks on the loaded data, looking for orphaned items, duplicate names, relations pointing to retired assets and mandatory fields left empty. Ask the vendor or your administrator which checks are built in and which need configuration or scripting. Done means you have the count and type of inconsistencies found in each instance and the time needed to resolve them.
Measure the manual work each model needs
Log every manual action taken in steps three to five, including data cleanup, custom field creation, plugin configuration and report building, and classify each as one-time setup or recurring upkeep. Compare the totals for each platform and note who performed each task and what skills it required. Done means you have a per-platform effort log that separates initial modeling from the work needed to keep the CMDB accurate each month.
Review results against the impact questions
Return to the list from step one and score how many questions each platform answered directly, with partial answers and gaps noted. Share the results with the change manager, the CMDB owner and the security lead, and agree which platform fits your change and impact needs, and which gaps need a quote-backed scope. Done means a signed-off record of answers, gaps and open questions for each platform, ready to inform a written implementation scope.
What running each CMDB costs in effort and licensing
The license fee is the smallest line in most of these budgets. iTop is AGPL-3.0 and free to download, and Combodo offers iTop Community for up to 10 agents, with paid Atelier for up to 50 agents and Forge for more than 50, either on premises or as SaaS. GLPI is GPL-3.0-or-later with no license fee for the community release. Teclib' sells GLPI Network subscriptions on top of that: the GLPI Network subscriptions page lists public cloud at 19 euros per user per month, private cloud at 21 euros per user per month with a 25-user minimum, and self-hosted tiers from 100 euros per month (Basic, up to 10 agents and 500 assets) to 4,500 euros per month (Enterprise). Confirm current figures on the vendor pages before budgeting, because the tier you need depends on your agent and asset counts, not on the feature list alone.
Most of the cost is one-time and recurring effort, and it is paid to people. The one-time work is data modeling: in iTop, defining classes, lifecycles, relations and impact paths for your CMDB, plus importing existing records through CSV. In GLPI, it is setting up assets, the GLPI Agent rollout and any GLPI Inventory plugin needed for network discovery. Recurring work is upgrades. The GLPI update procedure requires backing up the database, the files and the config directory including glpi.key and glpicrypt.key, then running php bin/console db:update. Since GLPI 11.0, plugins are suspended after a new version is detected and must be reactivated after major updates, so every plugin you rely on becomes a regression test. iTop's extensions come from the iTop Hub, so ask whether each one has been tested against the version you plan to run.
When you request a quote, ask for the work to be split by component: data model and migration, agent or discovery rollout, integrations with your monitoring and change tools, upgrade and backup runbooks, and handover of documentation. Ask whether support is included, whether it is commercial support from Combodo or Teclib', and at what level, since GLPI Network tiers differ in level 3 support. Ask the vendor or partner what it assumes about instance count, agent count and asset count, and request a written statement of what changes if your agent count crosses a tier boundary. DeployOpen plans and implements these engagements on a fixed scope agreed in writing, so the scope document should name each of those items explicitly.
Questions buyers ask before choosing a CMDB
Can GLPI do impact analysis the way iTop does?
GLPI covers asset-centric inventory, ITIL tickets, problems and changes, and contracts, but the supplied facts do not describe a graphical impact analysis feature comparable to iTop's. Ask Teclib' or test in a sandbox whether GLPI's relationships between assets, services and changes answer the questions your change board asks, such as which services fail when a database host is patched. iTop's README describes graphical impact analysis as a core feature, so the test for that need should be run on iTop first.
Does iTop need a custom build before it is useful?
iTop ships with a customizable CMDB, helpdesk, change and SLA management, but the data model is meant to be shaped to your estate. Expect some configuration work before the first useful report: deciding which classes, attributes and lifecycle states match your services and changes. Teams that accept the default model and only load data can start faster, while teams with unusual service maps should budget for model design. Confirm with Combodo or a partner which extensions from the iTop Hub you need.
How much effort does a custom CMDB model take to build and keep?
Effort depends on the number of classes you add, how many integrations feed them, and who owns the model after go-live. Build effort is one-time, but a model needs upkeep as services change, so the recurring cost is the time your administrators spend on schema changes, upgrades and data quality. Ask for an estimate that separates model design, data import, integration and training, and ask who will maintain the model once the project closes.
What does the GLPI Agent cover, and what needs the GLPI Inventory plugin?
The GLPI Agent collects hardware and software inventory from Windows, macOS and Linux endpoints and sends it to GLPI, and it is the successor of FusionInventory Agent. Per the GLPI documentation, native inventory through the agent has been available since version 10. Network discovery and SNMP inventory require the separate GLPI Inventory plugin. If your CMDB depends on network devices, plan that plugin into the scope and confirm it against your GLPI version before you commit.
Can a team run GLPI or iTop without outside help?
A capable team can run either platform, but the answer depends on skills and on the scope of the first release. Both need a web server, PHP and a supported database, and both need routine backup and upgrade discipline. GLPI's update procedure requires backing up the database, files and config directory including the glpi.key and glpicrypt.key files, and plugins are suspended after a new version is detected. iTop 3.3.x requires PHP 8.2 to 8.4 and GraphViz. Outside help is most useful for model design, impact analysis rules and cutover planning.
What drives the cost of a GLPI or iTop CMDB project up or down?
Costs fall into software, subscriptions, implementation and recurring operations. GLPI Network subscriptions are priced per user for cloud and per tier for self-hosted, with self-hosted plans from 100 euros per month for Basic to 4,500 euros per month for Enterprise, per the GLPI Network subscriptions page. iTop's free Community edition covers up to 10 agents, while Atelier covers up to 50 agents and Forge covers more than 50, per combodo.com. The main drivers are agent count, number of data classes, integration work and the staff hours you spend on model design. Ask every vendor or partner for a written quote that lists those items separately.
Which platform is safer to choose if we are unsure about change management?
Both platforms include change management, so the risk lies in how well each model matches your process. iTop's change management sits on a CMDB whose relationships you define, which suits teams that want change impact tied to services. GLPI's ITIL changes are tied to its asset inventory and contracts, which suits teams whose main goal is accurate asset records. Run a pilot change with your real approval path on both before you decide, and confirm supported workflows with each vendor.
How DeployOpen helps evaluate and implement a CMDB
DeployOpen plans, implements, integrates and documents self-hosted platforms inside your own infrastructure, so the CMDB choice is tested against your real change, discovery and asset workflows before anything is built. Engagements are fixed-scope, with scope, deliverables and timelines agreed in writing before work starts.
How we scope
- Map which processes need the CMDB first: change approval, incident impact, or asset tracking.
- Inventory the asset classes, relationships and discovery sources your written scope must cover.
- Agree in writing which GLPI or iTop editions, agent counts and plugins the design requires.
How we build
- Configure the GLPI asset model or the iTop data model and relationship classes to match your scope.
- Connect discovery, inventory agents and authentication sources such as LDAP to the chosen platform.
- Test impact analysis, change workflows and data import against sample records before go-live.
How we hand over
- Document the data model, integrations, configuration decisions and the upgrade path for your team.
- Deliver a written runbook covering backups, plugin behavior during upgrades and agent deployment.
- Walk your administrators through daily CMDB maintenance, data consistency checks and change records.
How we support
- Offer ongoing support, upgrades and continuous testing as a retainer after handover.
- Review each new release of the platform for schema, plugin and requirement changes before you upgrade.
- Confirm with the vendor which support tier fits your environment before you commit to a plan.
