A CMDB is only as good
as whatever fills it.
The Certero ServiceNow Application synchronises the enriched inventory from the platform into your CMDB. Windows, macOS and Linux computers arrive with their disks, network adapters, memory modules, volumes, software and software installations. SNMP network devices arrive on the same run.
Answered on one platform. One data model across ITAM, SAM, SaaS, Cloud and AI.
CMDB sync · scheduled run
IllustrativeCertero ServiceNow Application
Staging cleared, datasources queried, transforms run
- Staging tables cleared before the first query
- done
- Windows computers disks, adapters, memory, volumes
- staged
- macOS and Linux computers the same record types, the same run
- staged
- SNMP network devices their own datasource
- staged
- Transforms into the CMDB one transform map per datasource
- running
History record
pending → running → completed
ServiceNow Partner
Certero inventory synchronised into the ServiceNow CMDB
The problem
A CMDB does not go wrong. It goes stale.
It was right the week somebody populated it. Then the machines changed, the record did not, and the service desk started working from a document rather than from the thing itself.
- 01 typed once, true once
The record was accurate on the day a person created it.
Since then the machine has taken more memory, lost a disk and gained nine applications. Nobody was ever going to keep up by hand, and the record has been quietly drifting away from the hardware ever since.
Native inventory agent for Windows, macOS, Linux, AIX, HP-UX and Solaris ITAM
- 02 a scan sees what answers
The laptop was asleep, so as far as the CMDB is concerned it does not exist.
A discovery sweep records what replies while it is running. The travelling laptop, the locked-down segment and the machine nothing has credentials for are all absent, and an absence in a CMDB reads exactly like a machine you do not own.
Agentless and command-line inventory (csinvcli) for locked-down environments ITAM
- 03 a name and nothing under it
The CI exists. Open it and there is nothing inside.
No disks, no adapters, no memory, no file systems, no software. A technician takes the incident and goes looking for the detail somewhere else, which is the moment the CMDB stopped being the place people check.
Disks, network adapters, memory modules and volumes as default datasources ITAM
- 04 two records, one machine
The rebuild came back as a second CI, and the first one never left.
Duplicates and long-dead records are what a hand-fed CMDB accumulates. Every one of them makes the next person trust the next record a little less.
Duplicate system detection and stale device archiving ITAM
- None of this is fixed by another clean-up project. It is fixed by giving the CMDB a feed and letting the clean-up be the run.
Two of the five disciplines.
Each runs standalone and shares one asset model, so the parts of this that span products need no integration work.
From the machine to the CI, in five moves.
The application does one job and does it on a schedule: read the platform API, stage what comes back, transform it into the CMDB, and keep the record of what it did.
-
Fill the record before it leaves the platform
Whatever the CMDB is going to know, the platform has to see first. Ten discovery methods land in one schema, so the machines a network sweep never reaches are in the same inventory as the ones it does.
- Network Discovery across NetBIOS, SNMP and ICMP ITAM
- Native inventory agent for Windows, macOS, Linux, AIX, HP-UX and Solaris ITAM
- Agentless and command-line inventory (csinvcli) for locked-down environments ITAM
- Standalone inventory for air-gapped and offline systems ITAM
- Duplicate system detection and stale device archiving ITAM
- Software Recognition Database with centrally maintained title categorisation SAM
-
Install the application, set three things
API Url, API Username and API Password are all it needs to run. Page size and history retention are there when you want to tune them, and the application works out of the box otherwise.
- Certero ServiceNow Application, installed into your instance ServiceNow
- Three settings to start: API Url, API Username, API Password ServiceNow
- API Page Size and History Retention, both configurable ServiceNow
- Read-only Certero API as the single source the sync reads ServiceNow
- Menu access restricted to the x_certero.administrator role ServiceNow
-
Run it on a schedule, or run it now
The schedule is a ServiceNow scheduled script execution, running the sync in the background. Sync Now covers the times you need one before the next is due.
- Scheduled script execution triggering background sync ServiceNow
- Sync Now on the datasources and history pages ServiceNow
-
Stage it, then transform it into the CMDB
Each run begins by pruning history past the retention you set. Staging tables are then cleared, each datasource queries the API and pages until it is exhausted, and each transform map writes its records into ServiceNow.
- Default datasources for Windows, macOS and Linux computers ServiceNow
- Disks written as CMDB storage devices, volumes as CMDB file systems ServiceNow
- Network adapters, memory modules, software as CMDB packages, and software installations ServiceNow
- SNMP network devices on their own datasource ServiceNow
- A transform map created for every default datasource ServiceNow
-
Keep the record of every run
Each sync writes a history record that moves from pending to running to completed, with a system-message log underneath it, kept for as long as the retention you set.
- History record per sync: pending, running, completed ServiceNow
- System-message log against each run for debugging ServiceNow
- Old history pruned to the configured retention ServiceNow
- Datasource metadata held in a table, extendable to further API endpoints ServiceNow
Ask to see any one of these running in the product itself, on the screen where it happens.
A configuration item should arrive with its parts.
The application ships a datasource for Windows, macOS and Linux computers, and for each of those the disks, network adapters, memory modules, volumes, software and software installations. Every default datasource has a transform map that writes it into the CMDB.
In CerteroX
One machine, inventoried once.
Whichever of the ten discovery methods reached it, the record comes out the same shape — the machine, and everything inside it.
- Windows
- macOS
- Linux
Each of the three ships as seven datasources, one for each of the seven record types.
In ServiceNow
Seven records, one datasource and one transform map each.
- The computer itself
- Disks CMDB storage devices
- Network adapters
- Memory modules
- Volumes CMDB file systems
- Software CMDB packages
- Software installations
SNMP network devices arrive on a datasource of their own, and datasource metadata is held in a table — so further API endpoints can be added without changing the application.
What good looks like.
The test of a CMDB is whether anybody checks it before they pick up the phone.
- 01
The CI arrives with its parts.
Disks as storage devices, volumes as file systems, software as packages, plus network adapters, memory modules and software installations. The record answers the question that made someone open it.
- 02
Nothing depends on anyone remembering.
The scheduled script execution runs the sync in the background. The CMDB moves when the inventory moves, not when a clean-up gets scheduled.
- 03
Windows is not the only citizen.
macOS and Linux ship as default datasources on the same footing as Windows, with the same record types under each machine, and SNMP network devices come across on their own datasource.
- 04
Every run can be accounted for.
A history record per sync, pending through running to completed, with a system-message log for anything that needs explaining and a retention setting that stops the history growing forever.
- 05
A new endpoint does not need a new application.
Datasource metadata lives in a table, so the sync can be pointed at further API endpoints without changing the application itself.
Questions worth asking us.
Not the one you came with? Ask it directly and we will answer it in writing.
Does the application discover anything by itself?
No, and it is worth being exact about it. Discovery and inventory happen in CerteroX, across ten methods and six operating system families. The application reads the platform API and writes configuration items into the CMDB. It is a feed, not a second discovery tool competing with the one you have.
How much configuration is there before the first sync?
Three settings: API Url, API Username and API Password. The application ships its default datasources and their transform maps, so it works from there with minimal configuration. API page size and history retention are the two you may want to tune afterwards.
What actually arrives in the CMDB?
Windows, macOS and Linux computers, and for each of those the disks, network adapters, memory modules, volumes, software and software installations. Disks land as CMDB storage devices, volumes as CMDB file systems and software as CMDB packages. SNMP network devices come across on a datasource of their own.
How often does it sync, and can we force one?
A ServiceNow scheduled script execution triggers the sync in the background on whatever schedule you set. Sync Now sits on the datasources page and on the history page for the times you need one immediately.
How do we tell what a run did?
Every sync writes a history record that moves from pending to running to completed, with a system-message log against it for debugging. Old records are pruned to the retention you configure, so the history stays useful rather than becoming its own storage problem.
We want data the default datasources do not carry. Is that a customisation?
Not to the application. Datasource metadata is held in a table, so further API endpoints can be added as datasources without changing the application code.
Does the licence position move into ServiceNow too?
No. The effective licence position stays in CerteroX SAM, where the entitlement, the usage evidence and the publisher engines are. What crosses into ServiceNow is inventory: the configuration items, and the detail underneath them.
Open a CI.
See what arrived under it.
A sync run walked end to end: staging cleared, each datasource paged out of the API, each transform map writing into the CMDB, and the history record it leaves behind. Tell us which CMDB classes matter most to your service desk and we open those first.
Nothing to connect, and someone on the call who can answer for the datasources and the transform maps.