CSV Import
A guided 5-step ETL wizard for importing data from any CSV file into the Clarity CMDB. Supports custom column mapping, data transformation rules, and intelligent reconciliation against existing records — ideal when live API integrations aren't available.
Overview
The CSV Import module provides a guided 5-step ETL (Extract, Transform, Load) wizard for bringing data from any CSV file into the Clarity CMDB. Unlike the live API integrations for VMware, Nutanix, and cloud platforms, CSV Import works with static files — making it the right tool whenever a live connection isn't available or appropriate.
The wizard walks you through uploading your file, mapping each column to a CMDB field, applying transformation rules to clean the data, reviewing how the CSV rows match against existing records, and finally running the import. Each stage gives you visibility and control over what gets created, updated, or skipped.
When to Use CSV Import
Prerequisites
CSV Import requires no integration endpoints or credentials. You only need a well-formed CSV file and knowledge of which CMDB fields you want to populate.
- A CSV file with a header row. The first row must contain column headers — the wizard uses these to display the mapping interface. Blank header cells will be ignored.
- Known delimiter. The wizard defaults to comma-separated. If your file uses a different delimiter (tab, semicolon, pipe), you can specify it during upload.
- Target asset type identified. Know which CMDB asset type you're importing into (Device, Application, Database, Network, etc.) before starting the wizard.
Step 1 — Upload File
Navigate to Integrations → CSV Import → New Import to open the wizard.
Click Upload and select your .csv file from the file browser. Files up to 50 MB are supported.
Choose the CMDB asset type you're importing into from the Target Asset Type dropdown (e.g. Device, Application, Database). This determines which CMDB fields are available in the column mapping step.
If your file uses a non-standard delimiter, expand Advanced Options and set the delimiter character. The preview grid will update to confirm the file is parsed correctly.
Confirm the file preview looks correct, then click Next to proceed to Column Mapping.
Step 2 — Column Mapping
The wizard displays each column header from your CSV alongside a dropdown for selecting the matching CMDB field.
For each CSV column, select the corresponding CMDB field from the dropdown. Clarity will attempt auto-mapping for common column names (e.g. Name, IP Address, OS).
Leave unused columns set to Ignore — they will not be imported.
Click Save Mapping As… and give the template a descriptive name (e.g. RVTools VM Export Template or Legacy CMDB Device Export). Saved templates appear in a dropdown on the mapping step for future imports — saving significant time on recurring imports.
Click Next to proceed to the Transformation step.
Step 3 — Column Transformations
Transformation rules let you clean and normalise data before it is written to the CMDB. Rules are optional — you can skip this step if your data is already clean.
For any mapped column, click + Add Rule. Select the transformation type:
- Find & Replace — replace exact text (e.g. replace Windows Server 2019 Standard with Windows Server 2019).
- Upper / Lowercase — normalise text case.
- Prefix / Suffix — prepend or append static text to every value.
- Regex Replace — advanced pattern-based replacement for complex normalisation.
Rules run in priority order — lower number runs first. Set priorities when you have multiple rules on the same column and their order matters (e.g. run Find & Replace before Lowercase).
The preview grid shows transformed values in real time. When the output looks correct, click Next to proceed to Reconciliation.
Step 4 — Reconciliation
Reconciliation is where Clarity compares your CSV rows against existing CMDB records to determine what should be inserted, updated, or flagged for review. This step prevents duplicate records and keeps existing data clean.
Conflicts represent rows where Clarity found more than one possible match in the CMDB. Approving conflicts without reviewing them is the most common cause of duplicate records. Take the time to manually resolve each conflict.
New Records are CSV rows that don't match any existing CMDB record. They will be inserted as new records on import. Review the list to confirm these are genuinely new assets, not mismapped duplicates.
Matched Records are CSV rows that confidently match a single existing CMDB record. On import, these rows will update the existing record with the values from the CSV. Review a sample to confirm matches are correct.
Conflicts are rows where Clarity found multiple possible matches (ambiguous). For each conflict, you can:
- Approve — manually select which existing record to update.
- Reject — skip this row entirely (it won't be inserted or updated).
- Create New — treat this row as a new record and insert it.
Once all conflicts are resolved, click Next to proceed to the final Import step.
Step 5 — Import
The final step shows a summary: N to insert, N to update, N to skip. Confirm the numbers look correct before proceeding.
Click Run Import. A progress bar shows real-time import status. Do not close the browser tab during import.
When complete, the results summary shows the count of records inserted, updated, and skipped. The run is also logged in Integrations → History for audit purposes.
Reconciliation Reference
The reconciliation step uses the following categories to classify each CSV row:
| Category | Description | Default Action |
|---|---|---|
| Match Field | The CMDB field used to match CSV rows to existing records (e.g. Name or IP Address). Set during column mapping. | Configured per import |
| New Count | Rows where no matching record was found in the CMDB. These will be inserted as new records. | Insert |
| Match Count | Rows confidently matched to exactly one existing CMDB record. These will update the existing record. | Update |
| Conflict Count | Rows matched to multiple existing records (ambiguous). Require manual resolution before import can proceed. | Requires manual review |
Example Workflow
A project manager has an Excel spreadsheet of 200 servers exported from a legacy CMDB. The spreadsheet has five columns: ServerName, IPAddress, OS, BusinessUnit, and Owner. This data isn't available from the VMware integration — it lives only in the legacy system.
She exports the spreadsheet to CSV and opens the CSV Import wizard. In the mapping step, she maps ServerName → Name, IPAddress → IP Address, OS → Operating System, BusinessUnit → Business Unit, and Owner → Owner. She adds a Find & Replace transformation on the OS column to normalise names (e.g. "Windows Server 2019 Standard" → "Windows Server 2019") and saves the mapping as "Legacy CMDB Device Export".
The reconciliation step shows 180 matched records (matched to existing VMware-imported VMs by name), 20 new records (servers not yet in the CMDB), and 0 conflicts. She approves all and runs the import. The CMDB is now enriched with Business Unit and Owner data for all 200 servers — fields the VMware integration doesn't provide.
Tips
Most CSV imports are recurring — the same source system exports data in the same format every time. Saving the mapping template means you can re-run the same import in seconds next time rather than re-mapping every column from scratch.
- Use transformations to normalise OS names. Inconsistent OS naming (e.g. "Win 2019", "Windows Server 2019 Standard", "Windows 2019") creates multiple OS values in the CMDB. Use Find & Replace rules in the transformation step to standardise them before import.
- Use IP Address as the match field when names are inconsistent. If your CSV rows and existing CMDB records use different naming conventions, IP Address is often a more reliable match key than Name.
- Review the Conflicts category carefully. Conflicts are almost always where duplicate records lurk. A few minutes spent reviewing them prevents hours of data cleanup later.
- Don't import the full spreadsheet — filter first. If your spreadsheet contains retired or decommissioned assets, filter them out before exporting to CSV. Importing them creates unnecessary CMDB noise.
Common Mistakes & Troubleshooting
- Not reviewing the Conflicts category. Approving all conflicts without manually resolving them is the single most common cause of duplicate records in the CMDB. Always inspect each conflict and make a deliberate choice: update, skip, or insert as new.
- Mapping the wrong column to the Name field. If you accidentally map a column like Description or Asset Tag to Name, every record in the CMDB will be updated with the wrong name. Double-check your mappings — especially the Name field — before running the import.
- Not saving the mapping template. If you invest 10 minutes mapping 20 columns and don't save the template, you'll repeat that work on the next import. Always save the template before clicking Next.
- Running the import without checking the reconciliation summary. The reconciliation summary gives you a final sanity check before data is written. If the New Count is unexpectedly high (suggesting matches aren't working) or Conflict Count is large, investigate before proceeding.
- Importing a file with merged cells or non-standard formatting. CSV files exported from Excel with merged header cells or multi-row headers will not parse correctly. Ensure the CSV has a single, clean header row before uploading.