Resolving Process Historian PHRDY Buffer Overflow and Missing Tag Logging Trends in PCS 7

In SIMATIC PCS 7 systems with Process Historian (PH 2014) and Information Server, missing trend lines in Tag Logging coupled with active OS alarms 'PHRDY: server buffer limit reached' or 'buffer overflow' occur when the OS Server's PH Ready service cannot commit historical archive packets to Process Historian. This bottleneck is caused by unmapped Text Library language entries, exhausted SQL disk storage on the PH server, or network communication drops across Microsoft Message Queuing (MSMQ). To resolve this, align language mappings in the OS Text Library, expand PH database storage, and clear the local PH Ready swap buffer cache.

Architecture of the PH Ready Synchronization Pipeline

In SIMATIC PCS 7 architectures utilizing SIMATIC Process Historian (6ES7652-7AX58-0YB0), Operator Stations (OS Servers 6ES7652-0XX08-1XB0) continuously transfer short-term circular archive segments into the centralized historical database.

This synchronization relies on the PH Ready service (PH_Ready.exe):
1. Local Segment Generation: WinCC Tag Logging and Alarm Logging write real-time process data into local Microsoft SQL Server partition files.
2. Buffer Spooling: The PH Ready service intercepts closed segments and serializes them into local swap files on the OS Server hard drive (typically located in D:\WinCC_Projects\...\PH_Ready\Swap\).
3. Network Transport: Spooled packets are dispatched over the terminal bus via Microsoft Message Queuing (MSMQ) to the Process Historian server.
4. Historian Commit: The Process Historian database engine verifies schema integrity, uncompresses data records, and inserts them into enterprise time-series tables.

Root Causes of PHRDY Buffer Overflow and Empty Tag Logging Graphs

When the synchronization pipeline breaks, the OS Server issues two critical diagnostic alarms: PHRDY: server buffer limit reached and PHRDY: buffer overflow. Simultaneously, historical trend graphs in Process Historian and Information Server flatline or fail to display values.

Failure FactorUnderlying Failure MechanismObserved Diagnostic SymptomLanguage Text MismatchOS project contains technological messages in languages unmapped in PH.PH Management Console: Language text not mapped; SQL insert aborts.PH Database FullProcess Historian SQL database or transaction log volume out of disk space.MSMQ rejects incoming packets; OS buffer directory expands rapidly.MSMQ Network Queue BlockedFirewall blocks TCP port 1801 or network latency causes RPC timeout.OS outgoing message queue fills; PH_Ready.log reports connection drop.License OverrunConfigured archive tags exceed installed Process Historian tag licenses.PH refuses ingestion of un-licensed tags; tags omitted from graphs.

Table 1: Primary failure modes in Process Historian data ingestion.

Resolving Text Library Language Mismatches and MSMQ Stalls

A frequent, subtle cause of ingestion failure is unmapped multilingual text:
1. Open the OS project in SIMATIC WinCC Explorer.
2. Open the Text Library editor.
3. Check the configured languages: ensure that all defined languages (e.g., English-US, German) have corresponding translated column entries.
4. If a block or alarm message contains blank text in the default language, the Process Historian XML parser encounters an unhandled null exception during metadata synchronization.
5. In WinCC Explorer, verify that the Default Runtime Language in Computer Properties matches the regional language settings configured on the Process Historian server.

Managing the OS Server Local Buffer Cache and Queue Rollover

When the buffer limit alarm fires, the OS Server local disk is at risk of filling up:

Verify Process Historian Server Health:
* Log into the Process Historian server.
* Open Windows Disk Management: verify that SQL data and log drives have adequate free capacity (> 15% free space).
* Open Process Historian Management Console: confirm that the server status is Ready and database status is Online.
Inspect MSMQ Outgoing Queues:
* On the OS Server, open Computer Management (compmgmt.msc).
* Expand Services and Applications > Message Queuing > Outgoing Queues.
* Locate the destination queue pointing to the Process Historian IP address.
* If state shows Waiting to connect, verify network routing and execute SIMATIC Security Controller to reopen firewall rules.
Clear Corrupted Swap Files:
* If PH_Ready.log indicates a corrupted cache segment, stop the SIMATIC Process Historian Ready service.
* Navigate to the swap folder (...\PH_Ready\Swap), isolate the corrupted segment, and restart the service.
* The service will resume transferring buffered records, clearing the buffer alarm.

Verification

To confirm proper restoration of data archiving:

On the OS Server, open the WinCC Alarm Control: verify that alarms PHRDY: server buffer limit reached and buffer overflow disappear or acknowledge cleanly.
In the OS Server diagnosis directory, inspect PH_Ready.log: confirm status logs Transmission succeeded and buffer utilization drops below 20%.
In Process Historian Management Console, open the Diagnostics view: verify that incoming Tag Logging telegram rates return to normal values.
Open the Information Server web portal: load a 24-hour trend for an archived analog tag and confirm continuous, uninterrupted data curves.

Limitations and Operational Precautions

Never delete the entire PH_Ready\Swap directory while the PH Ready service is running, as active file handles can corrupt the synchronization registry.
Ensure adequate disk space is reserved on OS Servers for the swap buffer (at least 20 GB dedicated) to prevent buffer overflows during extended PH server maintenance.
Process Historian and Information Server require valid runtime licenses (6ES7652-7AX58-0YB0, 6AV6372-2DF07-0AX0). Always validate licensing packages with an authorized Siemens distributor.

Engineer FAQ

How long can the OS Server buffer data during a Process Historian outage?

Buffer retention time depends on allocated disk space and tag logging frequency. With a typical 20 GB buffer allocation and standard PCS 7 plant archiving rates, the OS Server can buffer data for 48 to 72 hours without data loss.

Why did client graphs show moving axes but no data points?

The trend control client updates its time axis based on the local computer clock, giving the illusion of activity, while the data channel receives null records from the desynchronized database.

Can I relocate the PH Ready swap folder to a different hard drive?

Yes. In the WinCC project directory, edit the PHReady.ini configuration file to point SwapPath to a high-capacity non-system drive (e.g., D:\PH_Swap\).