Diagnosing Process Historian IPC Hardware Sizing, Memory Starvation, and Storage Threshold Failures

In SIMATIC Process Historian (PH), a condition where segment backups fail, the PH tray icon indicates 'No connection to Process Historian', and client applications report 'Unexpected error occurred connecting to remoting server proxy' is typically caused by storage partition starvation or Microsoft SQL Server RAM exhaustion. When available storage drops below the mandatory 15% safety threshold, PH halts automated segmentation and enters protective Maintenance or Locked mode, terminating the .NET Remoting proxy. To resolve this, audit physical drive capacities across OS, database, and backup partitions, configure SQL Server maximum memory caps, ensure the backup destination share grants full permissions to NT SERVICE\MSSQL$HISTORIAN, and restart the PH services.

Hardware Sizing and Safety Threshold Architecture in Process Historian

SIMATIC Process Historian (PH) is engineered as an enterprise-grade, high-throughput archiving repository designed to collect millions of continuous process variables, alarms, and batch records from distributed SIMATIC PCS 7 OS Servers. Because database operations require high I/O operations per second (IOPS), Siemens establishes strict hardware configuration baselines: multi-core enterprise CPUs (Intel Xeon), a minimum of 16 to 32 GB RAM, and physically separated RAID arrays for the operating system (RAID 1), database files (RAID 5/10), transaction logs (RAID 1), and backup targets.

To ensure database integrity against sudden out-of-disk crashes, Process Historian continuously polls drive health. If free space on any monitored partition drops below 15% or 10 GB, the PH supervisory engine executes an emergency lockdown: new segment creation is blocked, segment transfers are halted, and the system transitions into Locked mode.

Root Causes: Storage Starvation and Microsoft SQL Server RAM Exhaustion

When Process Historian exhibits sudden connection failures, two primary hardware resource bottlenecks are responsible:

Partition Storage Exhaustion: If the drive housing the active database (HISTORIAN_Data.mdf) or transaction log (HISTORIAN_Log.ldf) fills up due to uncommitted transaction accumulation, PH immediately halts all background archiving jobs.
Uncapped SQL Server Memory Consumption: By default, Microsoft SQL Server dynamically allocates all available operating system RAM. Without a configured memory cap, SQL Server can starve Windows of physical memory, forcing critical background processes into high-latency paging files. When the CCProcessHistorianServer.exe process cannot obtain memory threads within internal timeout windows, its internal IPC proxies crash.

The Internal .NET Remoting Proxy and IPC Communication Breakdown

The error message 'Unexpected error occurred connecting to remoting server proxy' originates from the inter-process communication layer:

Remoting Architecture: The Process Historian Management Console (PH-MC), WinCC OS synchronization agents, and SIMATIC Information Server communicate with the PH core engine via Microsoft .NET Remoting / Windows Communication Foundation (WCF) TCP endpoints (ports 5000–5010).
Proxy Severance: If the underlying SQL Server service stalls or PH enters Locked mode, the remoting listener thread terminates. Client consoles attempting to connect receive an unhandled proxy communication exception, and the tray icon switches to red indicating 'No connection'.

Auditing Backup Destination Share and NTFS Security Credentials

Segment backups cannot execute if target network shares (NAS/SAN) reject transfer credentials:

Service Account Verification: The automated backup job is executed by the Windows service SQL Server (HISTORIAN). This service runs under the local virtual account NT SERVICE\MSSQL$HISTORIAN or a designated Active Directory domain account.
Share and File Permissions: The destination folder must possess explicit Full Control share permissions and NTFS Modify/Write permissions for this service account.
Path Validation: In PH-MC, verify that the backup destination uses a fully qualified UNC path (\\Storage_Server\PH_Backups\) rather than a transient mapped Windows drive letter (e.g. Z:\), which is not visible to non-interactive background system services.

Diagnostic Matrix: Process Historian Hardware and Service Faults

The following matrix outlines troubleshooting steps for common Process Historian infrastructure failures:

Diagnostic SymptomUnderlying Root CauseCorrective Engineering RemedyTray icon red; 'No connection'Free disk space below 15% thresholdExpand partition or delete obsolete temp files; verify >20% free'Remoting server proxy' errorPH background service crashed or deadlockedRestart SIMATIC Process Historian Server service via services.mscSegment backup status 'Failed'Target UNC share lacks NTFS write rightsGrant Full Control to MSSQL$HISTORIAN or domain service accountSQL Server consumes 98% RAMSQL Max Server Memory unconstrainedOpen SSMS, set 'Max Server Memory' to 75% of physical RAMTransaction log (.ldf) swellingFull recovery model without log backupsSwitch recovery model to Simple or configure scheduled log backups

Step-by-Step Restoration and Service Recovery Procedure

Follow this sequence to restore Process Historian to an active state:

Free Storage Space: Open Windows Explorer on the PH Server. Ensure that all drives (C:, Data, Log, Backup) have at least 20% free space.
Configure SQL Memory Ceiling: Open SQL Server Management Studio (SSMS), right-click the server instance HISTORIAN, select Properties -> Memory, and cap maximum server memory to 75% of physical RAM (e.g., 24,000 MB on a 32 GB server).
Verify Windows Services: In services.msc, verify that SQL Server (HISTORIAN) and SQL Server Agent (HISTORIAN) are Running.
Restart PH Service: Right-click SIMATIC Process Historian Server and select Restart.
Reset Segment Error State: Open PH-MC, navigate to Database -> Segments, right-click any faulted segment, and select Retry Backup.
Confirm that the system state returns to Active and the tray icon turns green.

Verification

Verify the full recovery of Process Historian through the following checks:

Confirm that the PH system tray icon displays a green checkmark indicating 'Process Historian: Active'.
In PH-MC, verify that the Dashboard status reports 'System State: Active' and all disk volume bars display green.
Trigger an on-demand segment backup; confirm that .mdf and .ldf archive chunks appear in the target network backup folder.
In SIMATIC Information Server, run a historical trend report to verify that data retrieval across the remoting interface executes cleanly.

Limitations and Operational Precautions

Adhere to the following operational safeguards:

Never Modify SQL Tables Directly: Do not edit Process Historian SQL database tables directly using custom SQL scripts; schema modifications can permanently invalidate database integrity and void manufacturer support.
Dedicated Server Isolation: Process Historian must run on dedicated physical or virtual server hardware; never co-locate PH with WinCC OS Servers, Domain Controllers, or Engineering Stations.
RAID Monitoring: Ensure hardware RAID controllers are monitored via vendor utilities (e.g., Dell OpenManage) to detect failing physical disks before array degradation interrupts database I/O.

Engineer FAQ

Can mapped drive letters (e.g. Z:) be used for PH segment backups?

No. Background Windows services running under virtual service accounts cannot access interactive user drive mappings. Always use fully qualified UNC paths (\Server\Share).

Does restarting the PH server cause data loss on WinCC OS Servers?

No. WinCC OS Servers retain process data locally in their revolving ArchiveManager queues during temporary PH offline periods and forward the backlog automatically upon reconnection.

What should I do if the transaction log file (HISTORIAN_Log.ldf) grows to hundreds of gigabytes?

Execute a full database backup in SQL Server Management Studio, truncate the transaction log, and verify that scheduled SQL maintenance plans are active.