Werk #18116: Improve storage mechanism for BI aggregations
Component | BI | ||||
Title | Improve storage mechanism for BI aggregations | ||||
Date | May 19, 2025 | ||||
Level | Trivial Change | ||||
Class | Bug Fix | ||||
Compatibility | Incompatible - Manual interaction might be required | ||||
Checkmk versions & editions |
|
Prior to this change, the compiled and frozen aggregations were stored to disk using the aggregation ID and branch title as file names. This was problematic for some systems since the file name was truncated after a certain length. This led to missing aggregations when running service discovery.
To fix this, we are now using a hashing algorithm to store the compiled and frozen aggregations. This ensures that the file name length will never exceed the file system's limit.
In addition, a migration script will run during update that will clear the compiled aggregations and migrate the legacy frozen aggregations to the new format.