Need to know more in detail what the function "Rebuild Index" does.

If you go to the properties of the Main and Auto Summary partitions you can use the button: "Rebuild Index". But there isn't any explanation to find what exactly happens after you press this button. The help only tells you this: "Select Rebuild Index to manually rebuild the index"

After pressing the button, the content of the folder "C:\Historian\Data\DataIndex\IndexCircular" will be wiped. After 5 to 10 minutes, you will see appearing some History blocks in there.

Who can explain me what exactly happens. I don't believe the Runtime database is touched when using this rebuild function, so it has nothing to do with a rebuild of any indexes on tables.
When to use? Only when you cannot retrieve values for particular tags? Or is ir recommended to use this on a regular base?

Parents
  • For every AVEVA Historian storage shard (a.k.a. partition, or history timeline = a sequence of history blocks) the storage process maintains a mapping of a tag-version to the history blocks where data values may be present. That mapping gets stored in the corresponding DataIndex folder and helps storage to achieve faster data retrieval across long time durations. In some rare situations that DataIndex may go out of sync with its history timeline resulting in not seeing values from some history blocks. You can force the index rebuilding manually by shutting down / disabling historian and deleting the DataIndex folder manually. Then start the historian and give it some time (depending on the amount of stored data) to recreate that index. During that time some queries may be slower than usual. So that button is just a user-friendly way to do the same thing without disabling the historian server.  Use it only if you have reasons to believe that the data for a tag is not visible for some history blocks. You do not need to run it on a regular base.

Reply
  • For every AVEVA Historian storage shard (a.k.a. partition, or history timeline = a sequence of history blocks) the storage process maintains a mapping of a tag-version to the history blocks where data values may be present. That mapping gets stored in the corresponding DataIndex folder and helps storage to achieve faster data retrieval across long time durations. In some rare situations that DataIndex may go out of sync with its history timeline resulting in not seeing values from some history blocks. You can force the index rebuilding manually by shutting down / disabling historian and deleting the DataIndex folder manually. Then start the historian and give it some time (depending on the amount of stored data) to recreate that index. During that time some queries may be slower than usual. So that button is just a user-friendly way to do the same thing without disabling the historian server.  Use it only if you have reasons to believe that the data for a tag is not visible for some history blocks. You do not need to run it on a regular base.

Children
No Data