Configuring PI Data Archive to Overwrite Old Archives: A Practical Guide for Limited Storage Environments

In development and testing environments, it's common to face disk space constraints—especially when running a PI Data Archive server with limited drive capacity for archive files. For many, the default behavior of PI Data Archive is to create new archive files as older ones fill up, which can eventually fill the disk and halt operations until manual intervention occurs. Fortunately, the PI Data Archive includes tuning parameters that let you control archive behavior, including instructing the system to overwrite old archives when needed.

Understanding PI Archive Behavior

By default, the PI Data Archive will automatically create new archive files as the current (primary) archive reaches full capacity. This is done using the Archive_AutoArchiveFileRoot tuning parameter, which specifies the path and file name prefix for new archives. While this is ideal for production environments where data retention is essential, it can quickly fill up available storage on development or test systems.

How to Configure Overwriting of Old Archives

To switch from automatic archive file creation to a circular, overwrite behavior, follow these key steps:

1. Remove the Archive_AutoArchiveFileRoot Parameter

This parameter controls the automatic creation of new archive files. You can adjust this in PI System Management Tools (SMT):

• Navigate to Operation > Tuning Parameters in PI SMT.
• Locate the Archive_AutoArchiveFileRoot parameter and remove it.

Once removed, PI Data Archive will no longer create new archive files as the primary archive fills. Instead, when the primary archive reaches capacity, it will shift—the oldest archive will be re-purposed as the new primary and new data will overwrite its contents.

2. Set Archive_OverwriteDataOnAutoShiftFailure to 1

This parameter ensures that, if the server cannot create a new archive (for example, due to the disk being full), it will overwrite the data in the oldest archive:

• In PI SMT, locate the Archive_OverwriteDataOnAutoShiftFailure parameter.
• Set its value to 1.

This forces the archive subsystem to recycle the oldest archive file and overwrite its data, ensuring that data collection continues even when space is limited.

3. Monitor Archive Locations

The location of archive files is determined at creation, but the overwrite behavior is not path-dependent. If your archives reside on various drives, PI Data Archive will simply recycle the oldest file, regardless of its path. However, maintaining consistent directory structures can simplify administration.

Important Caveats

• Data Loss: This overwrite strategy means that older data will be permanently lost as it is replaced by newer information. Always ensure that you are comfortable with this data retention policy before implementing it outside of non-production environments.
• No Automatic Expansion: With Archive_AutoArchiveFileRoot removed, no new archive files will be created. Ensure you have the desired number of archives available for rotation.
• Monitor Archive Sizing: Properly size your archive files for the right data window, since this will determine your effective data retention period.

Summary Table

Conclusion

Enabling archive overwriting on PI Data Archive is a handy technique for dev/test environments with tight disk space constraints. By removing automatic new archive creation and enabling recycling via tuning parameters, you can keep your system running without manual intervention—even when storage is scarce. Always remember to monitor data retention needs and apply this strategy judiciously; it is not recommended for production systems where historical data preservation is critical.

If you have experiences or insights on PI Data Archive sizing and management, feel free to share them in the comments!