Resolving Backup Size Discrepancies (800 GB to 3 GB) and Storage Management in SIMATIC Process Historian
When backing up a SIMATIC Process Historian database whose active storage consumed 800 GB of hard disk space and observing that the resulting backup file occupies only 3 GB, the discrepancy is caused by confusion between raw historical segment archives and relational database schema metadata. The 3 GB file represents either the Microsoft SQL Server configuration and master database catalog or a single recent segment backup, rather than the complete historical Process Historian archive. In Process Historian, raw historical tags and alarms are segmented across individual database segment files (*.mdf / *.ldf). To execute a complete backup and free disk space, engineers must configure automated long-term segment backups via the Process Historian Management Console or SQL Server Maintenance Plans, verify segment checksums, and purge detached segments.
Table of Contents
- Executive Summary
- Process Historian Database Architecture: Master Database vs Segmented Tag Archives
- Why Database Backups Shrink from 800 GB to 3 GB: The Metadata Fallacy
- Verifying Historical Archive Data Integrity Before Deletion
- Step-by-Step Procedure for Complete Process Historian Backup and Disk Recovery
- Process Historian Database Storage and Backup Strategy Matrix
- Verification
- Purging and Deleting Historical Segments to Free Hard Disk Space
- Limitations and Operational Precautions
- Engineer FAQ
- Technical References
- Related Articles
- Related Tags
Executive Summary
In large SIMATIC PCS 7 process plants, the SIMATIC Process Historian (PH) serves as the central high-capacity long-term data repository, continuously archiving millions of process values, alarms, and batch tracking records from multiple OS Server pairs. As operational years elapse, Process Historian storage volumes expand, frequently consuming 800 GB to several terabytes across physical hard disk arrays. When hard disk space nears exhaustion, system administrators follow database backup documentation to export the database to external storage. However, upon completing the backup, administrators are alarmed to discover that the generated backup file is only 3 GB in size. Fearing catastrophic data loss, engineers hesitate to delete any files. This technical guide explains the internal database structure of Process Historian, identifies why the 3 GB file was generated, and outlines safe data evacuation procedures.
Process Historian Database Architecture: Master Database vs Segmented Tag Archives
Understanding the Process Historian SQL database topology is essential to evaluating backup archives:
- The System Catalog Database (
HISTORIAN_STORAGE): Holds relational metadata, tag lists, compression algorithms, server connections, and user rights. In an 800 GB installation, this metadata database typically occupies 2 GB to 5 GB. - Historical Segment Databases (
PH_TAG_*,PH_ALARM_*): The actual historical process tags and alarms are partitioned into discrete single database segments (*.mdf/*.ldf) spanning specific time intervals (e.g. 1 month per segment).
* The Core Insight: Performing a standard SQL Server database backup onHISTORIAN_STORAGEbacks up only the 3 GB metadata catalog, completely omitting the hundreds of gigabytes of historical segment files stored on dedicated archive storage partitions.
Why Database Backups Shrink from 800 GB to 3 GB: The Metadata Fallacy
The dramatic size reduction from 800 GB to 3 GB occurs because of how backup tools interact with SQL Server:
- The administrator executed a backup script or SQL Server Management Studio (SSMS) task targeting exclusively the active master database container.
- The active segment container contains pointers to historical segments, but the physical segment files reside in dedicated archive folders (e.g.
D:\ProcessHistorian\ArchiveStore\). - The 3 GB backup file is completely valid for database configuration and schema, but contains zero historical trend samples from previous years.
Verifying Historical Archive Data Integrity Before Deletion
Before deleting or purging any data on an 800 GB Process Historian server:
- Audit Physical Archive Files: Open Windows Explorer and navigate to the configured archive directories. Confirm the presence of historical segment database files (
*.mdf/*.ldf). - Check Segment Status in PH Management Console:
* Launch the Process Historian Management Console.
* Navigate toDatabase -> Segments.
* Check the operational state of each segment:Active,Attached,Backed Up, orDetached. - Inspect Information Server Reports: Generate an Information Server report querying historical values across multiple years. If data displays normally, the historical segments are intact in the storage pool.
Step-by-Step Procedure for Complete Process Historian Backup and Disk Recovery
To execute a true full backup of all 800 GB of historical records and safely free disk space:
- Configure Automated Segment Backup via PH Console:
* In the Process Historian Management Console, navigate toConfiguration -> Backup.
* Enable Automatic Backup.
* Specify a secondary high-capacity backup destination (e.g. an external RAID array, NAS share, or SAN LUN).
* The Process Historian background service will systematically detach older segments, write verified backup images (*.bak/*.mdf), and verify archive checksums. - Perform Full Storage Database Backup:
* In SQL Server Management Studio (SSMS), perform a full backup ofHISTORIAN_STORAGE(producing the 3 GB metadata file). - Execute Segment Detachment:
* In the PH Management Console, select verified backed-up segments older than the desired online retention window (e.g. older than 1 year).
* ExecuteDetach Segment. - Relocate or Purge Detached Segment Files:
* Once a segment is detached in the console, its corresponding MDF/LDF files are unlocked by SQL Server.
* Move these files to external enterprise backup storage, or delete them if verified copies exist on secondary media.
* Hard disk space on the Process Historian server will immediately recover by hundreds of gigabytes.
Process Historian Database Storage and Backup Strategy Matrix
The following table summarizes Process Historian storage components and backup methodologies:
HISTORIAN_STORAGE).mdf, .ldf, .bak.mdf, .ldf.mdf files.mdf filesVerification
Confirm that Process Historian data is safely backed up and storage recovered:
- Backup Volume Verification: Confirm that the total byte volume on the backup destination matches the anticipated volume (~800 GB of segment files plus the 3 GB metadata file).
- Disk Free Space Audit: On the Process Historian server, open Windows Explorer. Confirm that local drive
C:orD:indicates substantial free storage (e.g. > 500 GB free). - Segment Re-attachment Test: In the PH Management Console, select one detached historical segment from the backup drive and click
Attach Segment. Verify that Information Server can query data from that re-attached segment without database errors.
Purging and Deleting Historical Segments to Free Hard Disk Space
When freeing disk space on Process Historian:
- Never Delete Active Segments: Never delete the currently active segment (status
ActiveorCurrent). Deleting active files crashes the Process Historian acquisition service. - Use Management Console Detachment: Always detach segments via the official console before moving physical files; never delete files directly from Windows Explorer while SQL Server holds active locks.
Limitations and Operational Precautions
Never perform standard third-party file compression or Windows NTFS compression on Process Historian storage drives. Compressing active SQL database partitions causes severe transaction log corruption and disk latency.
When expanding Process Historian storage hardware (SAS RAID controllers, SAN storage, or Archive Tag licenses), all hardware and software specifications must be validated against official Siemens catalogs and confirmed through an authorized Siemens distributor quotation.
Engineer FAQ
Can I delete the 3 GB backup file?
No. The 3 GB backup file contains the essential master catalog and database schema required to restore the entire system during disaster recovery.
How long does Process Historian take to back up 800 GB?
Backup times depend on network bandwidth and storage write speeds; backing up 800 GB over a Gigabit network typically requires 4 to 8 hours.
What happens if hard disk space reaches 100% on Process Historian?
When disk space is exhausted, SQL Server halts write transactions. The Process Historian stops acquiring data, and local OS Servers buffer data in temporary caches until full.
Can detached segments be searched in Information Server?
No. Detached segments must be re-attached via the PH Management Console before Information Server can query their historical records.