Queries on tier-2 Historian do not seem to use auto summary data

Disclaimer: I don't know for sure since it's been a while since I was tinkering with this feature myself. From what I recall, the tier-2 should behave just the same way as a tier-1 and not really be aware of the fact that it is used in a tier-2 capacity or act any differently because of its data source.

I think we need Elliott Middleton  to chime in here. 

Yes I'm sure tier-2 uses autosummary as well. And we have seen that in the autosummary circular files. But when we try to retrieve the data using those typical auto summay queries, we are not able to force retrieval from auto summary.

And on the opposite side, when we instruct the tier-1 to retrieve from main circular instead of autosummary, the query time is like the tier-2.

Rickard is correct, if we're talking about "simple" replicated tags (sends all the stored data to the "tier 2"). Auto-summary does not apply for "summary replication". 

How are you determining if auto-summary is used or not? Is it based on the response times or are you checking the "wwOptions" column in the query results? The column is the right way to reliably know the source.If that reports "primarydata", it is not using auto-summary. That could be for a variety of reasons, including:

1. The query doesn't match the auto-summary criteria (>1 hour interval, cyclic/best fit or aggregate mode, etc.)
2. Auto-summary is disabled in System Parameters
3. Too many tags in the "tier 2" (by default, only the first 70K tags have auto-summary enabled)

Or, of course, it might be a defect. If you suspect that, include version details and confirm 1-3 do not apply, and follow up with Tech Support. 
 

We are running the same query on tier1 and tier2 (only the tagname is changed since it has a prefix on tier2). Loading times on Historian web client are also increased on tier-2.

Auto-summary is enabled, and there are not so many tags.

The way we have used to determine if auto-summary i used is firstly by looking at execution times. If it takes 200 times as long to execute the query on tier-2, I think it is reasonable to assume it is summarizing the data on the fly and not using auto-summary. Like we mentioned previously, if we instruct the tier-1 to fetch from PRIMARYDATA, it gets roughly the same execution time as tier-2.

Secondly we have also looked at resource monitor to see which files are accessed when we run the query. We can clearly see that it opens up main circular-files, not auto-summary.

I ran a query at a customer with auto-summary disabled. And wwOption still says AUTOSUMMARYDATA even though auto-summary data does not exist and is disabled.

Thanks for clarifying and I agree with you that the "tier 2" isn't leveraging auto-summary like it should. I suggest following up with Tech Support and include the version details.

Thanks for clarifying and I agree with you that the "tier 2" isn't leveraging auto-summary like it should. I suggest following up with Tech Support and include the version details.