Hi, i have a datatable with 150.000 rows. I noticed that the time for querying rows is increasing by increasing rows number. Now is about 15 seconds to query 150.000 total row with 4 filters and result is about 15 rows.
In that table i have 3 primary keys(Date,MachineID,subID) I see that a service called "Reindex" needs of "index" in configuration inside datatable options. I've created this index on machineID and Date but when i test it with test button thingworx says:
Unable to Invoke Service Reindex on fermatedatatable_bck : com.thingworx.exceptions.NotImplementedException: Not Implemented
What does it mean ? Do my version of thingworx not support this function ? how can i reindex my datatable for increase query speed ?
If you have a datatable over 100k rows you are out of luck using standard DataTables and you better connect directly to a Database through the JDBC connector. Read this post: Where Should I Store My Thingworx Data?
I don't think this is the desired answer but actually this is how TW works, we should push them harder to improve this BIG limitation.
I installed TW on Postgres so all data inside datatable TW object are stored on a DataBase table. I thought postgres let me to use large table data for storing in datatable so as you said before it is a BIG limitation. But the topic of my post was the reindex :-) anyone else found this error please?
Yes you are right with the problem of the index/question, but still if you solve the index problem you will have the performance issue, using PostgreSQL for standard TW DataTable doesn't solve performance issue.
I'm now on TW 7.3 and PostgreSQL testing and yes it throws this exception on DataTables, I've opened a Case. But anyway, I think Indexes are working out of the box with PostgreSQL, it was needed on Neo4J but it seems it's not needed on PostgreSQL anymore.
How are you trying to use Indexes querying?
Hi Alberto, extending on what Carles have already mentioned a documentation SPR has been created to improve the logging for this exception, which it is not as the service is intended for Neo4j. You can access the document attached for improving the logging using this article