Repository navigation
Conversation
|
For comparison's sake, here are the plans for executing via Jena: Notice how the filter operator is under the multijoin vs. it being over it in direct HeFQUIN. |
|
The requests printed out for direct HeFQUIN execution: Jena execution: Notice this extra request in the direct HeFQUIN execution: |
|
Thanks for the detailed information about the issue that you have run into with this change. Two more questions: |
It should be the last query performed right? In that case, it should be this request: If I run the query with --skipExecution, the cardinality requests are made without causing a Too Many Requests so I want to say the error occurs during query execution. |
|
Regarding 2, notice that your print out mentions two requests. Assuming it is the second one of these two that causes the error, can you try to do just this query with curl (here is a comment with some curl commands that you can adapt: #737 (comment)) If that works without problems, can you put the pattern of that query into a single SERVICE clause and execute that as a query with HeFQUIN. Also, assuming you get the Too Many Requests error as an exception printed after execution, how does the complete stack trace of this exception look like? |
It seems to work and return results when using curl.
Running this query: also worked without problems. The full stack trace is this: |
|
Could it be that the extra request that occurs when executing the query directly via HeFQUIN is enough to cause the Too Many Requests? |
It seems so. I mean, it is the only possible reason that I can see at the moment. The strange thing is that the other endpoint does not complain about too many requests, even if HeFQUIN sends three requests to that one as well. And another thing that is strange is that, if you run the query again, the complaint doesn't come already for the first of the three request. Or, perhaps it does!? If you look at the "query planning info" of the Wikidata request ( Can you please check what exactly is happening with the cardinality requests for this query by putting a debug break point into that if-block (or temporarily putting a |
|
I printed out the following in the aforementioned if-block: Some fields are evidently more helpful than others, can't say this has clarified anything for me though. |
|
If |
|
It seems to be this: |
Bypass Jena execution when HeFQUIN can directly execute the query.
WIP at the moment, notably
ExampleQuery1.rqconsistently returns429with source assignment:and logical plan:
Notes before making it PR:
HeFQUINQuerySupportCheckerandResultSetSolutionMappingneed to be moved to a more appropriate placeHeFQUINEngine